Appendix 10-E: Step-by-Step Configuration Procedures¶
Table of Contents¶
SECTION 1: PRE-IMPLEMENTATION CHECKLIST¶
- 1.1 Access & Permissions Verification
- 1.2 Licensing Verification
- 1.3 Network & Connectivity Requirements
SECTION 2: WEBEX CONTROL HUB - AI AGENT CONFIGURATION¶
- 2.1 Creating a Webex AI Agent
- 2.2 Configuring Intents
-
2.3 Configuring Responses
Deferred — See AI Implementation Chapter
Response template configuration (TTS settings, fallback messages, language variants) is covered in 10.3.4 Response Templates & TTS Settings. Detailed click-by-click steps will be added once the Webex AI Agent UI is confirmed stable post-GA.
SECTION 3: GOOGLE CLOUD PLATFORM - PROJECT SETUP¶
- 3.1 Creating a GCP Project
- 3.2 Enabling Required APIs
- 3.3 Creating a Service Account
-
3.4 Configuring IAM Permissions
Deferred — Pending GCP Project Setup
IAM role assignments for the Dialogflow CX service account require the GCP project to be created first (3.1). Minimum required roles:
roles/dialogflow.adminon the project,roles/iam.serviceAccountTokenCreatorfor the webhook service account. Full procedure will be drafted after GCP project provisioning is confirmed.
SECTION 4: DIALOGFLOW CX - AGENT CREATION¶
- 4.1 Creating a Dialogflow CX Agent
- 4.2 Configuring Agent Settings
- 4.3 Creating Intents
-
4.4 Creating Entities
Deferred — See NLU Design Chapter
Entity definitions (PRODUCT_TYPE, ACCOUNT_TYPE, LANGUAGE, CITY) are specified in 10.6 NLU & Intent Design. Step-by-step Dialogflow CX console procedure will be added in Phase 2B implementation sprint.
-
4.5 Building Flows
Deferred — See Conversational Flows Chapter
Flow architecture and design patterns are in 10.7 Conversational Flows. Console build procedure will be added after flow designs are signed off in UAT.
-
4.6 Creating Pages
Deferred — Linked to Flow Build
Pages are created as part of flow building (4.5 above). Procedure will be included in the same sprint.
-
4.8 Setting Up Environments
Deferred — Pre-Production Gate
Dialogflow CX environments (draft, staging, production) are configured after intent and flow development is complete. Required: draft environment promoted to staging after UAT sign-off, production promotion gated on Phase 2B go-live approval. Procedure will be documented at that stage.
SECTION 5: WEBEX CONTACT CENTER - CCAI CONNECTOR¶
- 5.1 Creating CCAI Configuration
-
5.2 Linking Dialogflow CX Agent
Deferred — Pending CCAI Connector Availability
The Webex Contact Center AI (CCAI) connector for Dialogflow CX requires the GCP project and Dialogflow agent to be fully configured first. Linking procedure: Control Hub > Contact Center > Integrations > Google CCAI > Add Integration. Full step-by-step will be added once the connector is activated in the tenant.
-
5.3 Configuring Voice Settings
Deferred — TTS Provider Selection Pending
Voice settings (TTS engine, voice profile, language, speaking rate) depend on the TTS provider decision (Google TTS vs Cisco TTS). Decision is tracked in Deferred Items Register. Procedure will be drafted once provider is confirmed.
SECTION 6: FLOW DESIGNER - VIRTUAL AGENT INTEGRATION¶
- 6.1 Adding Virtual Agent V2 Node
-
6.2 Configuring Node Parameters
Deferred — Flow Designer GA Required
Node parameter configuration varies by Flow Designer release. Parameters for Virtual Agent, HTTP Request, and Queue Contact nodes are documented in 6.5 IVR Flow Implementation. Appendix procedure will align once the production tenant is provisioned.
-
6.4 Passing Context Variables
Deferred — See Webhook Integration Chapter
Context variable passing between Flow Designer and Dialogflow CX (via CCAI connector parameters) is specified in 10.8 Webhook & Fulfillment. Implementation procedure will be added post-UAT.
SECTION 7: AGENT ASSIST CONFIGURATION¶
- 7.1 Enabling Agent Assist
- 7.2 Configuring Knowledge Base
- 7.3 Setting Up Suggestions
- 7.4 Configuring Conversation Summaries
SECTION 8: VALIDATION & TESTING¶
1.1 Access & Permissions Verification¶
Before beginning configuration, verify all required access and permissions are in place.
Procedure: Verify Webex Control Hub Access¶
+-----------------------------------------------------------------------------+
| PROCEDURE 1.1.1: VERIFY CONTROL HUB ADMINISTRATOR ACCESS |
+-----------------------------------------------------------------------------+
| |
| Step 1: Open browser and navigate to: |
| https://admin.webex.com |
| |
| Step 2: Log in with your administrator credentials |
| Username: [your-admin-email]@abhavtech.com |
| Password: [your-password] |
| |
| Step 3: After login, verify your role: |
| a. Click your profile icon (top-right corner) |
| b. Click "My Profile" |
| c. Scroll to "Administrative Roles" section |
| |
| Step 4: Confirm you have one of these roles: |
| [ ] Full Administrator |
| [ ] Contact Center Administrator |
| |
| [!]️ CHECKPOINT: |
| Expected: "Contact Center Administrator" or higher role visible |
| If missing: Contact your Webex organization administrator |
| |
| Step 5: Verify Contact Center access: |
| a. Click "Services" in left navigation panel |
| b. Verify "Contact Center" appears in the services list |
| c. Click "Contact Center" |
| d. Verify dashboard loads without errors |
| |
| [OK] VERIFICATION COMPLETE when Contact Center dashboard displays |
| |
+-----------------------------------------------------------------------------+
Procedure: Verify GCP Access¶
+-----------------------------------------------------------------------------+
| PROCEDURE 1.1.2: VERIFY GOOGLE CLOUD PLATFORM ACCESS |
+-----------------------------------------------------------------------------+
| |
| Step 1: Open browser and navigate to: |
| https://console.cloud.google.com |
| |
| Step 2: Log in with your GCP credentials |
| Account: [your-email]@abhavtech.com |
| (Use Google Workspace or Cloud Identity account) |
| |
| Step 3: Verify organization access: |
| a. Click the project dropdown (top navigation bar) |
| b. Click "ALL" tab |
| c. Verify "abhavtech.com" organization is visible |
| |
| Step 4: Verify required IAM roles: |
| a. Navigate to: IAM & Admin -> IAM |
| b. Find your email in the members list |
| c. Confirm you have these roles (or equivalent): |
| [ ] Project Creator |
| [ ] Billing Account User |
| [ ] Dialogflow API Admin |
| |
| [!]️ CHECKPOINT: |
| Expected: All three roles visible for your account |
| If missing: Contact your GCP organization administrator |
| |
| [OK] VERIFICATION COMPLETE when all roles confirmed |
| |
+-----------------------------------------------------------------------------+
1.2 Licensing Verification¶
Procedure: Verify WxCC Licenses¶
+-----------------------------------------------------------------------------+
| PROCEDURE 1.2.1: VERIFY WEBEX CONTACT CENTER LICENSES |
+-----------------------------------------------------------------------------+
| |
| Step 1: In Control Hub, navigate to: |
| Services -> Contact Center -> Settings |
| |
| Step 2: Click "Licenses" tab |
| |
| Step 3: Document current license allocation: |
| |
| License Type | Entitled | Consumed | Available |
| ----------------------+----------+----------+------------- |
| Standard Agent | ________ | ________ | ________ |
| Premium Agent | ________ | ________ | ________ |
| Supervisor | ________ | ________ | ________ |
| |
| Step 4: Verify minimum requirements for AI features: |
| [ ] At least 1 Premium Agent license available |
| (Premium required for AI Agent features) |
| |
| [!]️ CHECKPOINT: |
| Expected: Premium Agent licenses > 0 available |
| If insufficient: Contact Cisco account team for licensing |
| |
| Step 5: Verify Virtual Agent entitlement: |
| a. Navigate to: Features -> Virtual Agent |
| b. Verify "Create Virtual Agent" button is enabled |
| c. If disabled, check subscription includes AI features |
| |
| [OK] VERIFICATION COMPLETE when Premium licenses available and |
| Virtual Agent feature accessible |
| |
+-----------------------------------------------------------------------------+
Procedure: Verify GCP Billing¶
+-----------------------------------------------------------------------------+
| PROCEDURE 1.2.2: VERIFY GCP BILLING ACCOUNT |
+-----------------------------------------------------------------------------+
| |
| Step 1: In GCP Console, navigate to: |
| Navigation Menu (☰) -> Billing |
| |
| Step 2: Verify billing account exists: |
| a. Look for billing account in the list |
| b. Note the billing account name: _______________________ |
| c. Note the billing account ID: _______________________ |
| |
| Step 3: Verify billing account is active: |
| a. Click on the billing account name |
| b. Verify status shows "Active" (green indicator) |
| c. Verify payment method is configured |
| |
| Step 4: Verify you have billing permissions: |
| a. Click "Account management" in left panel |
| b. Verify you can view billing details |
| c. Verify role: "Billing Account User" or higher |
| |
| [!]️ CHECKPOINT: |
| Expected: Active billing account with valid payment method |
| If missing: Contact finance team to set up GCP billing |
| |
| NOTE: Dialogflow CX pricing is consumption-based: |
| - ~$0.007 per text request |
| - ~$0.001 per second of audio (voice) |
| - Estimated monthly cost for Abhavtech: $500-1,500 |
| |
| [OK] VERIFICATION COMPLETE when billing account active and accessible |
| |
+-----------------------------------------------------------------------------+
1.3 Network & Connectivity Requirements¶
Procedure: Verify Network Connectivity¶
+-----------------------------------------------------------------------------+
| PROCEDURE 1.3.1: VERIFY NETWORK REQUIREMENTS |
+-----------------------------------------------------------------------------+
| |
| The following endpoints must be accessible from WxCC: |
| |
| DIALOGFLOW CX ENDPOINTS: |
| ------------------------------------------------------------------------- |
| | Endpoint | Port | Protocol | Purpose |
| |---------------------------------------|------|----------|--------------|
| | dialogflow.googleapis.com | 443 | HTTPS | API calls |
| | asia-south1-dialogflow.googleapis.com | 443 | HTTPS | Regional API |
| | storage.googleapis.com | 443 | HTTPS | Audio files |
| | oauth2.googleapis.com | 443 | HTTPS | Auth |
| |
| WEBHOOK ENDPOINTS (if self-hosted): |
| ------------------------------------------------------------------------- |
| | Endpoint | Port | Protocol | Purpose |
| |---------------------------------------|------|----------|--------------|
| | api.abhavtech.com | 443 | HTTPS | Fulfillment |
| |
| Step 1: Verify firewall rules allow outbound HTTPS (443) to: |
| [ ] *.googleapis.com |
| [ ] *.google.com |
| |
| Step 2: If using corporate proxy, ensure: |
| [ ] Proxy supports HTTPS inspection bypass for Google domains |
| [ ] Certificate pinning is not blocking Google APIs |
| |
| Step 3: Test connectivity (from admin workstation): |
| Open command prompt/terminal and run: |
| |
| curl -I https://dialogflow.googleapis.com |
| |
| Expected response: HTTP/2 404 (API requires auth, but |
| connectivity confirmed) |
| |
| [!]️ CHECKPOINT: |
| Expected: Successful HTTPS connection to Google APIs |
| If blocked: Work with network team to allow Google API endpoints |
| |
| [OK] VERIFICATION COMPLETE when connectivity test successful |
| |
+-----------------------------------------------------------------------------+
2.1 Creating a Webex AI Agent¶
Procedure: Create New Virtual Agent¶
+-----------------------------------------------------------------------------+
| PROCEDURE 2.1.1: CREATE WEBEX AI AGENT |
+-----------------------------------------------------------------------------+
| |
| Estimated Time: 15 minutes |
| Prerequisites: Procedures 1.1.1 and 1.2.1 completed successfully |
| |
| ======================================================================= |
| |
| Step 1: Navigate to Virtual Agent Management |
| ------------------------------------------------------------------------- |
| a. Log into Control Hub: https://admin.webex.com |
| b. Click "Services" in left navigation |
| c. Click "Contact Center" |
| d. In left panel, click "Features" |
| e. Click "Virtual Agent" |
| |
| [Screenshot placeholder: Control Hub - Virtual Agent list] |
| |
| Step 2: Initiate Agent Creation |
| ------------------------------------------------------------------------- |
| a. Click "+ Create Virtual Agent" button (top-right) |
| b. Creation wizard opens |
| |
| Step 3: Select Agent Type |
| ------------------------------------------------------------------------- |
| a. In the "Agent Type" selection screen: |
| O Scripted |
| * Scripted with NLU <- SELECT THIS |
| O Dialogflow CX |
| O Dialogflow ES |
| b. Click "Next" |
| |
| Step 4: Configure Basic Information |
| ------------------------------------------------------------------------- |
| a. Enter the following values: |
| |
| Field | Value |
| ---------------------+---------------------------------------- |
| Name | Abhi_Simple_VA |
| Description | Simple voice IVR agent for basic |
| | intents - hours, callbacks, surveys |
| |
| b. Click "Next" |
| |
| Step 5: Configure Language Settings |
| ------------------------------------------------------------------------- |
| a. Primary Language: Select "English (United States) - en-US" |
| b. Additional Languages: Leave unchecked for now |
| (Hindi will be handled by Dialogflow CX) |
| c. Click "Next" |
| |
| Step 6: Configure Voice Settings |
| ------------------------------------------------------------------------- |
| a. Voice Selection: |
| - Click dropdown under "English (US)" |
| - Select "en-US-Neural2-J" (Male voice) |
| - Click [>] preview button to hear sample |
| |
| b. Voice Parameters: |
| Field | Value |
| ---------------------+---------------------------------------- |
| Speaking Rate | 1.0 (default) |
| Pitch | 0 (default) |
| |
| c. Click "Next" |
| |
| Step 7: Configure Conversation Settings |
| ------------------------------------------------------------------------- |
| a. Enter the following values: |
| |
| Field | Value | Notes |
| --------------------------+--------+-------------------------- |
| No-Input Timeout | 5 | seconds |
| Speech Complete Timeout | 2 | seconds |
| Max No-Input Retries | 3 | before escalation |
| Max No-Match Retries | 3 | before escalation |
| Barge-In | * On | allow interruption |
| DTMF Collection | * On | accept keypad input |
| Inter-digit Timeout | 3 | seconds |
| |
| b. Click "Next" |
| |
| Step 8: Configure Advanced Settings |
| ------------------------------------------------------------------------- |
| a. Enter the following values: |
| |
| Field | Value |
| --------------------------+---------------------------------- |
| Confidence Threshold | 0.6 |
| Logging Level | Standard |
| Session Timeout | 300 (seconds) |
| Enable Analytics | [x] Checked |
| |
| b. Click "Create" |
| |
| Step 9: Verify Creation |
| ------------------------------------------------------------------------- |
| a. Wait for "Virtual Agent created successfully" message |
| b. Verify "Abhi_Simple_VA" appears in the Virtual Agent list |
| c. Status should show "Draft" (no intents configured yet) |
| |
| [!]️ CHECKPOINT: |
| Expected: Agent "Abhi_Simple_VA" visible in list with "Draft" status |
| If error: Note error message and check prerequisites |
| |
| [OK] PROCEDURE COMPLETE - Continue to Procedure 2.2.1 |
| |
+-----------------------------------------------------------------------------+
2.2 Configuring Intents¶
Procedure: Create Intent - greeting.hello¶
+-----------------------------------------------------------------------------+
| PROCEDURE 2.2.1: CREATE INTENT - greeting.hello |
+-----------------------------------------------------------------------------+
| |
| Estimated Time: 10 minutes |
| Prerequisites: Procedure 2.1.1 completed |
| |
| ======================================================================= |
| |
| Step 1: Open Agent for Editing |
| ------------------------------------------------------------------------- |
| a. In Virtual Agent list, click "Abhi_Simple_VA" |
| b. Agent configuration page opens |
| c. Click "Intents" tab |
| |
| Step 2: Create New Intent |
| ------------------------------------------------------------------------- |
| a. Click "+ Add Intent" button |
| b. Intent creation form opens |
| |
| Step 3: Configure Intent Details |
| ------------------------------------------------------------------------- |
| a. Enter basic information: |
| |
| Field | Value |
| ---------------------+---------------------------------------- |
| Intent Name | greeting.hello |
| Description | Welcome and initial greeting |
| Priority | Normal |
| |
| Step 4: Add Training Phrases |
| ------------------------------------------------------------------------- |
| a. In "Training Phrases" section, click "+ Add Phrase" |
| b. Add each phrase below (press Enter after each): |
| |
| Training Phrase # | Text |
| ---------------------+---------------------------------------- |
| 1 | Hello |
| 2 | Hi |
| 3 | Hey |
| 4 | Good morning |
| 5 | Good afternoon |
| 6 | Good evening |
| 7 | Hi there |
| 8 | Hello there |
| 9 | Hey there |
| 10 | Greetings |
| 11 | I need help |
| 12 | Can you help me |
| 13 | I'm calling about |
| 14 | I have a question |
| 15 | Is anyone there |
| |
| c. Verify all 15 phrases appear in the list |
| |
| Step 5: Configure Response |
| ------------------------------------------------------------------------- |
| a. Scroll to "Response" section |
| b. Select Response Type: * Text-to-Speech |
| c. In the text box, enter: |
| |
| +-------------------------------------------------------------+ |
| | Hello! Welcome to Abhavtech. I'm Abhi, your virtual | |
| | assistant. | |
| | | |
| | I can help you with business hours and locations, request | |
| | a callback, or connect you to one of our team members. | |
| | | |
| | How can I help you today? | |
| +-------------------------------------------------------------+ |
| |
| Step 6: Configure Follow-up Action |
| ------------------------------------------------------------------------- |
| a. In "Follow-up Action" section: |
| * Listen for next input <- SELECT THIS |
| O End conversation |
| O Transfer to flow |
| |
| Step 7: Save Intent |
| ------------------------------------------------------------------------- |
| a. Click "Save" button |
| b. Wait for "Intent saved successfully" message |
| c. Intent appears in the intent list |
| |
| [!]️ CHECKPOINT: |
| Expected: "greeting.hello" intent visible with 15 training phrases |
| If error: Verify all required fields are completed |
| |
| [OK] PROCEDURE COMPLETE - Continue to Procedure 2.2.2 |
| |
+-----------------------------------------------------------------------------+
Procedure: Create Intent - hours.location¶
+-----------------------------------------------------------------------------+
| PROCEDURE 2.2.2: CREATE INTENT - hours.location |
+-----------------------------------------------------------------------------+
| |
| Estimated Time: 10 minutes |
| Prerequisites: Procedure 2.2.1 completed |
| |
| ======================================================================= |
| |
| Step 1: Create New Intent |
| ------------------------------------------------------------------------- |
| a. In Intents tab, click "+ Add Intent" |
| |
| Step 2: Configure Intent Details |
| ------------------------------------------------------------------------- |
| Field | Value |
| ---------------------+---------------------------------------- |
| Intent Name | hours.location |
| Description | Business hours and office location |
| Priority | Normal |
| |
| Step 3: Add Training Phrases (20 total) |
| ------------------------------------------------------------------------- |
| |
| Hours-related (10): |
| 1. What are your hours |
| 2. What time do you open |
| 3. What time do you close |
| 4. Are you open now |
| 5. Business hours |
| 6. When are you open |
| 7. Operating hours |
| 8. What are your working hours |
| 9. Are you open on weekends |
| 10. Are you open on Sunday |
| |
| Location-related (10): |
| 11. Where are you located |
| 12. What is your address |
| 13. Office location |
| 14. Where is your office |
| 15. How do I get to your office |
| 16. Directions to your office |
| 17. Office address |
| 18. Location please |
| 19. Where are you |
| 20. What city are you in |
| |
| Step 4: Configure Response |
| ------------------------------------------------------------------------- |
| Response Type: * Text-to-Speech |
| |
| +-------------------------------------------------------------+ |
| | Our main support center in Mumbai is available 24 hours | |
| | a day, 7 days a week. | |
| | | |
| | Our regional offices operate Monday through Friday: | |
| | India offices: 9 AM to 9 PM Indian Standard Time. | |
| | London office: 9 AM to 6 PM British Time. | |
| | New Jersey office: 9 AM to 6 PM Eastern Time. | |
| | | |
| | Our headquarters is located in Mumbai, India. | |
| | | |
| | Is there anything else I can help you with? | |
| +-------------------------------------------------------------+ |
| |
| Step 5: Configure Follow-up Action |
| ------------------------------------------------------------------------- |
| * Listen for next input |
| |
| Step 6: Mark as Contained |
| ------------------------------------------------------------------------- |
| a. Check: [x] Mark as contained (self-service complete) |
| (This flags the interaction for containment analytics) |
| |
| Step 7: Save Intent |
| ------------------------------------------------------------------------- |
| a. Click "Save" |
| |
| [!]️ CHECKPOINT: |
| Expected: "hours.location" intent with 20 phrases and containment |
| |
| [OK] PROCEDURE COMPLETE - Continue to Procedure 2.2.3 |
| |
+-----------------------------------------------------------------------------+
Procedure: Create Intent - callback.request¶
+-----------------------------------------------------------------------------+
| PROCEDURE 2.2.3: CREATE INTENT - callback.request |
+-----------------------------------------------------------------------------+
| |
| Estimated Time: 10 minutes |
| |
| ======================================================================= |
| |
| Step 1: Create New Intent |
| ------------------------------------------------------------------------- |
| Field | Value |
| ---------------------+---------------------------------------- |
| Intent Name | callback.request |
| Description | Customer requests a callback |
| Priority | High |
| |
| Step 2: Add Training Phrases (18 total) |
| ------------------------------------------------------------------------- |
| 1. Call me back |
| 2. Request a callback |
| 3. I want a callback |
| 4. Can you call me back |
| 5. Please call me back |
| 6. I'd like a callback |
| 7. Schedule a callback |
| 8. Callback please |
| 9. I don't want to wait |
| 10. Rather than wait can you call me |
| 11. Have someone call me |
| 12. Get someone to call me |
| 13. Don't want to hold |
| 14. Too long to wait |
| 15. I'll take a callback |
| 16. Put me on the callback list |
| 17. Call me instead |
| 18. Ring me back |
| |
| Step 3: Configure Response |
| ------------------------------------------------------------------------- |
| +-------------------------------------------------------------+ |
| | I'll arrange for one of our team members to call you back.| |
| | We'll call you at the number you're calling from. | |
| | Is that okay? | |
| +-------------------------------------------------------------+ |
| |
| Step 4: Configure Follow-up Action |
| ------------------------------------------------------------------------- |
| a. Select: O Transfer to flow |
| b. Flow Selection: Callback_Flow_v1 |
| c. Variables to pass: |
| - callback_requested = true |
| - callback_number = {{ANI}} |
| |
| Step 5: Mark as Contained |
| ------------------------------------------------------------------------- |
| [x] Mark as contained |
| |
| Step 6: Save Intent |
| ------------------------------------------------------------------------- |
| Click "Save" |
| |
| [OK] PROCEDURE COMPLETE - Continue to Procedure 2.2.4 |
| |
+-----------------------------------------------------------------------------+
Procedure: Create Intent - agent.handoff¶
+-----------------------------------------------------------------------------+
| PROCEDURE 2.2.4: CREATE INTENT - agent.handoff |
+-----------------------------------------------------------------------------+
| |
| Estimated Time: 10 minutes |
| |
| ======================================================================= |
| |
| Step 1: Create New Intent |
| ------------------------------------------------------------------------- |
| Field | Value |
| ---------------------+---------------------------------------- |
| Intent Name | agent.handoff |
| Description | Customer requests human agent |
| Priority | High |
| |
| Step 2: Add Training Phrases (25 total) |
| ------------------------------------------------------------------------- |
| |
| Direct requests: |
| 1. Speak to an agent |
| 2. Talk to a human |
| 3. Human please |
| 4. Agent please |
| 5. Real person |
| 6. Transfer me |
| 7. Connect me to someone |
| 8. I want to talk to a person |
| 9. Get me a representative |
| 10. Let me speak to someone |
| 11. Operator |
| 12. Representative |
| 13. Customer service |
| |
| Frustration indicators: |
| 14. I don't want to talk to a bot |
| 15. You're not helping |
| 16. This isn't working |
| 17. You don't understand |
| 18. I need a real person |
| 19. Stop, I want an agent |
| 20. Just transfer me |
| 21. Enough with the robot |
| 22. I'm done with this bot |
| |
| Polite variants: |
| 23. Could I please speak to someone |
| 24. Would it be possible to talk to a person |
| 25. May I speak with a representative |
| |
| Step 3: Configure Response |
| ------------------------------------------------------------------------- |
| +-------------------------------------------------------------+ |
| | Of course! I'll connect you with one of our team members | |
| | right away. Please hold while I transfer you. | |
| +-------------------------------------------------------------+ |
| |
| Step 4: Configure Follow-up Action |
| ------------------------------------------------------------------------- |
| a. Select: * Exit to flow with escalation |
| b. Exit Reason: agent_requested |
| c. Context Output Variables: |
| - escalation_reason = "customer_requested" |
| |
| Step 5: Save Intent |
| ------------------------------------------------------------------------- |
| Click "Save" |
| |
| [OK] PROCEDURE COMPLETE - Continue to Procedure 2.2.5 |
| |
+-----------------------------------------------------------------------------+
Procedure: Configure Fallback Intent¶
+-----------------------------------------------------------------------------+
| PROCEDURE 2.2.5: CONFIGURE FALLBACK INTENT |
+-----------------------------------------------------------------------------+
| |
| Estimated Time: 5 minutes |
| Note: Fallback is a system intent - it already exists |
| |
| ======================================================================= |
| |
| Step 1: Access Fallback Intent |
| ------------------------------------------------------------------------- |
| a. In Intents list, find "fallback.default" (system intent) |
| b. Click to edit |
| |
| Step 2: Configure Fallback Responses |
| ------------------------------------------------------------------------- |
| a. Response for Attempt 1: |
| +-------------------------------------------------------------+ |
| | I didn't quite catch that. Could you please repeat what | |
| | you need help with? You can say things like "business | |
| | hours" or "request a callback". | |
| +-------------------------------------------------------------+ |
| |
| b. Response for Attempt 2: |
| +-------------------------------------------------------------+ |
| | I'm still having trouble understanding. You can also | |
| | press 1 for Sales, 2 for Support, 3 for Billing, or 0 | |
| | to speak with an agent. | |
| +-------------------------------------------------------------+ |
| |
| c. Response for Attempt 3+ (Max retries): |
| +-------------------------------------------------------------+ |
| | Let me connect you with someone who can better assist you.| |
| | Please hold. | |
| +-------------------------------------------------------------+ |
| |
| Step 3: Configure DTMF Fallback Mapping |
| ------------------------------------------------------------------------- |
| a. Enable DTMF collection on fallback |
| b. Configure mappings: |
| |
| Key | Action | Flow Variable |
| ----+---------------------------+---------------------------- |
| 1 | Route to Sales | selected_queue = "Sales" |
| 2 | Route to Support | selected_queue = "Support" |
| 3 | Route to Billing | selected_queue = "Billing" |
| 0 | Route to Agent | selected_queue = "Support" |
| * | Repeat menu | (replay prompt) |
| |
| Step 4: Configure Max Retries Action |
| ------------------------------------------------------------------------- |
| After max retries: * Exit to flow with escalation |
| Exit Reason: max_retries_exceeded |
| |
| Step 5: Save Changes |
| ------------------------------------------------------------------------- |
| Click "Save" |
| |
| [OK] PROCEDURE COMPLETE - All intents configured |
| |
+-----------------------------------------------------------------------------+
2.4 Testing the Webex AI Agent¶
Procedure: Test Agent in Simulator¶
+-----------------------------------------------------------------------------+
| PROCEDURE 2.4.1: TEST WEBEX AI AGENT IN SIMULATOR |
+-----------------------------------------------------------------------------+
| |
| Estimated Time: 15 minutes |
| Prerequisites: All intents configured (Procedures 2.2.1-2.2.5) |
| |
| ======================================================================= |
| |
| Step 1: Access Simulator |
| ------------------------------------------------------------------------- |
| a. In Virtual Agent configuration page |
| b. Click "Test" button (top-right) |
| c. Simulator panel opens on right side |
| |
| Step 2: Test greeting.hello Intent |
| ------------------------------------------------------------------------- |
| a. In text input, type: Hello |
| b. Press Enter |
| c. Verify response matches expected greeting |
| |
| Expected Response: |
| "Hello! Welcome to Abhavtech. I'm Abhi..." |
| |
| [ ] PASS [ ] FAIL |
| |
| Step 3: Test hours.location Intent |
| ------------------------------------------------------------------------- |
| a. Type: What are your business hours |
| b. Press Enter |
| c. Verify hours/location response |
| |
| Expected Response: |
| "Our main support center in Mumbai..." |
| |
| [ ] PASS [ ] FAIL |
| |
| Step 4: Test callback.request Intent |
| ------------------------------------------------------------------------- |
| a. Type: I want a callback |
| b. Press Enter |
| c. Verify callback confirmation response |
| |
| Expected Response: |
| "I'll arrange for one of our team members..." |
| |
| [ ] PASS [ ] FAIL |
| |
| Step 5: Test agent.handoff Intent |
| ------------------------------------------------------------------------- |
| a. Click "Reset" to start new session |
| b. Type: I want to speak to a human |
| c. Press Enter |
| d. Verify handoff response |
| |
| Expected Response: |
| "Of course! I'll connect you..." |
| Exit Reason: agent_requested |
| |
| [ ] PASS [ ] FAIL |
| |
| Step 6: Test Fallback Behavior |
| ------------------------------------------------------------------------- |
| a. Click "Reset" |
| b. Type: asdfghjkl (gibberish) |
| c. Press Enter |
| d. Verify fallback response |
| |
| Expected Response: |
| "I didn't quite catch that..." |
| |
| [ ] PASS [ ] FAIL |
| |
| Step 7: Document Test Results |
| ------------------------------------------------------------------------- |
| |
| Test Summary: |
| | Intent | Status | Notes | |
| |---------------------|--------|----------------------------| |
| | greeting.hello | | | |
| | hours.location | | | |
| | callback.request | | | |
| | agent.handoff | | | |
| | fallback.default | | | |
| |
| [!]️ CHECKPOINT: |
| Expected: All 5 tests PASS |
| If failures: Review intent configuration and training phrases |
| |
| Step 8: Publish Agent |
| ------------------------------------------------------------------------- |
| a. Close simulator |
| b. Click "Publish" button |
| c. Confirm publication |
| d. Status changes from "Draft" to "Published" |
| |
| [OK] PROCEDURE COMPLETE - Webex AI Agent ready for flow integration |
| |
+-----------------------------------------------------------------------------+
3.1 Creating a GCP Project¶
Procedure: Create New GCP Project¶
+-----------------------------------------------------------------------------+
| PROCEDURE 3.1.1: CREATE GCP PROJECT FOR DIALOGFLOW CX |
+-----------------------------------------------------------------------------+
| |
| Estimated Time: 10 minutes |
| Prerequisites: Procedure 1.1.2 completed (GCP access verified) |
| |
| Project Details: |
| ------------------------------------------------------------------------- |
| Project Name: Abhavtech WxCC AI |
| Project ID: abhavtech-wxcc-ai |
| Region: asia-south1 (Mumbai) |
| Purpose: Dialogflow CX Virtual Agent |
| |
| ======================================================================= |
| |
| Step 1: Navigate to Project Creation |
| ------------------------------------------------------------------------- |
| a. Open browser: https://console.cloud.google.com |
| b. Log in with your GCP credentials |
| c. Click the project dropdown (top navigation bar) |
| d. Click "NEW PROJECT" button (top-right of popup) |
| |
| [Screenshot placeholder: GCP Console - New Project button] |
| |
| Step 2: Configure Project Details |
| ------------------------------------------------------------------------- |
| a. Project name: Abhavtech WxCC AI |
| b. Project ID: Click "EDIT" and enter: abhavtech-wxcc-ai |
| |
| [!]️ IMPORTANT: Project ID must be globally unique. |
| If "abhavtech-wxcc-ai" is taken, append a number: |
| e.g., abhavtech-wxcc-ai-001 |
| |
| c. Organization: Select "abhavtech.com" from dropdown |
| d. Location: Select appropriate folder or leave as organization |
| |
| Field | Value |
| -----------------+------------------------------------------ |
| Project name | Abhavtech WxCC AI |
| Project ID | abhavtech-wxcc-ai |
| Organization | abhavtech.com |
| Location | abhavtech.com (or specific folder) |
| |
| Step 3: Create Project |
| ------------------------------------------------------------------------- |
| a. Click "CREATE" button |
| b. Wait for project creation (30-60 seconds) |
| c. Notification appears: "Create Project: Abhavtech WxCC AI" |
| d. Click "SELECT PROJECT" in the notification |
| |
| Step 4: Link Billing Account |
| ------------------------------------------------------------------------- |
| a. After project loads, you may see "This project has no |
| billing account" warning |
| b. Click "Link a billing account" |
| c. Select your billing account from the dropdown |
| d. Click "SET ACCOUNT" |
| |
| OR navigate manually: |
| a. Click Navigation Menu (☰) |
| b. Click "Billing" |
| c. Click "LINK A BILLING ACCOUNT" |
| d. Select billing account |
| e. Click "SET ACCOUNT" |
| |
| Step 5: Verify Project Creation |
| ------------------------------------------------------------------------- |
| a. Click project dropdown (top navigation) |
| b. Verify "Abhavtech WxCC AI" appears in project list |
| c. Verify project ID shows as "abhavtech-wxcc-ai" |
| |
| [!]️ CHECKPOINT: |
| Expected: Project created with billing linked |
| Verify: Project dropdown shows "Abhavtech WxCC AI" |
| |
| Step 6: Record Project Information |
| ------------------------------------------------------------------------- |
| Document for future reference: |
| |
| Project Name: _______________________________ |
| Project ID: _______________________________ |
| Project Number: _______________________________ |
| (Find project number in Project settings) |
| |
| [OK] PROCEDURE COMPLETE - Continue to Procedure 3.2.1 |
| |
+-----------------------------------------------------------------------------+
3.2 Enabling Required APIs¶
Procedure: Enable Dialogflow CX API¶
+-----------------------------------------------------------------------------+
| PROCEDURE 3.2.1: ENABLE REQUIRED GOOGLE CLOUD APIS |
+-----------------------------------------------------------------------------+
| |
| Estimated Time: 5 minutes |
| Prerequisites: Procedure 3.1.1 completed |
| |
| Required APIs: |
| ------------------------------------------------------------------------- |
| 1. Dialogflow API |
| 2. Cloud Speech-to-Text API |
| 3. Cloud Text-to-Speech API |
| 4. Cloud Functions API (for webhooks) |
| 5. Cloud Run API (alternative for webhooks) |
| |
| ======================================================================= |
| |
| Step 1: Navigate to API Library |
| ------------------------------------------------------------------------- |
| a. Ensure "Abhavtech WxCC AI" project is selected |
| b. Click Navigation Menu (☰) |
| c. Navigate to: APIs & Services -> Library |
| |
| [Screenshot placeholder: GCP Console - API Library] |
| |
| Step 2: Enable Dialogflow API |
| ------------------------------------------------------------------------- |
| a. In search box, type: Dialogflow API |
| b. Click "Dialogflow API" in results |
| c. Click "ENABLE" button |
| d. Wait for API to enable (10-30 seconds) |
| e. "API enabled" confirmation appears |
| |
| [ ] Dialogflow API enabled |
| |
| Step 3: Enable Speech-to-Text API |
| ------------------------------------------------------------------------- |
| a. Click "APIs & Services" in breadcrumb to return to Library |
| b. Search for: Cloud Speech-to-Text API |
| c. Click the result |
| d. Click "ENABLE" |
| e. Wait for confirmation |
| |
| [ ] Cloud Speech-to-Text API enabled |
| |
| Step 4: Enable Text-to-Speech API |
| ------------------------------------------------------------------------- |
| a. Return to API Library |
| b. Search for: Cloud Text-to-Speech API |
| c. Click the result |
| d. Click "ENABLE" |
| e. Wait for confirmation |
| |
| [ ] Cloud Text-to-Speech API enabled |
| |
| Step 5: Enable Cloud Functions API |
| ------------------------------------------------------------------------- |
| a. Return to API Library |
| b. Search for: Cloud Functions API |
| c. Click the result |
| d. Click "ENABLE" |
| e. Wait for confirmation |
| |
| [ ] Cloud Functions API enabled |
| |
| Step 6: Enable Cloud Build API (required for Cloud Functions) |
| ------------------------------------------------------------------------- |
| a. Return to API Library |
| b. Search for: Cloud Build API |
| c. Click the result |
| d. Click "ENABLE" |
| |
| [ ] Cloud Build API enabled |
| |
| Step 7: Verify All APIs Enabled |
| ------------------------------------------------------------------------- |
| a. Navigate to: APIs & Services -> Enabled APIs & services |
| b. Verify all 5 APIs appear in the enabled list: |
| |
| [ ] Dialogflow API |
| [ ] Cloud Speech-to-Text API |
| [ ] Cloud Text-to-Speech API |
| [ ] Cloud Functions API |
| [ ] Cloud Build API |
| |
| [!]️ CHECKPOINT: |
| Expected: All 5 APIs show "Enabled" status |
| If any missing: Repeat enable steps for that API |
| |
| [OK] PROCEDURE COMPLETE - Continue to Procedure 3.3.1 |
| |
+-----------------------------------------------------------------------------+
3.3 Creating a Service Account¶
Procedure: Create Service Account for WxCC Integration¶
+-----------------------------------------------------------------------------+
| PROCEDURE 3.3.1: CREATE SERVICE ACCOUNT FOR CCAI CONNECTOR |
+-----------------------------------------------------------------------------+
| |
| Estimated Time: 10 minutes |
| Prerequisites: Procedure 3.2.1 completed |
| |
| Service Account Details: |
| ------------------------------------------------------------------------- |
| Name: WxCC CCAI Connector |
| ID: wxcc-ccai-connector |
| Purpose: Allow WxCC to authenticate with Dialogflow CX |
| |
| ======================================================================= |
| |
| Step 1: Navigate to Service Accounts |
| ------------------------------------------------------------------------- |
| a. Ensure "Abhavtech WxCC AI" project is selected |
| b. Click Navigation Menu (☰) |
| c. Navigate to: IAM & Admin -> Service Accounts |
| |
| [Screenshot placeholder: GCP Console - Service Accounts] |
| |
| Step 2: Create Service Account |
| ------------------------------------------------------------------------- |
| a. Click "+ CREATE SERVICE ACCOUNT" (top of page) |
| b. Service account creation wizard opens |
| |
| Step 3: Configure Service Account Details (Step 1 of 3) |
| ------------------------------------------------------------------------- |
| a. Enter the following: |
| |
| Field | Value |
| -------------------------+---------------------------------- |
| Service account name | WxCC CCAI Connector |
| Service account ID | wxcc-ccai-connector |
| | (auto-generated from name) |
| Description | Service account for Webex Contact |
| | Center CCAI integration |
| |
| b. Note the generated email address: |
| wxcc-ccai-connector@abhavtech-wxcc-ai.iam.gserviceaccount.com |
| |
| c. Click "CREATE AND CONTINUE" |
| |
| Step 4: Grant Service Account Access (Step 2 of 3) |
| ------------------------------------------------------------------------- |
| a. Click "Select a role" dropdown |
| b. Search for and select: "Dialogflow API Admin" |
| c. Click "+ ADD ANOTHER ROLE" |
| d. Search for and select: "Dialogflow API Client" |
| |
| Required Roles: |
| [ ] Dialogflow API Admin |
| [ ] Dialogflow API Client |
| |
| e. Click "CONTINUE" |
| |
| Step 5: Grant Users Access (Step 3 of 3) |
| ------------------------------------------------------------------------- |
| a. This step is optional for our use case |
| b. Click "DONE" |
| |
| Step 6: Verify Service Account Created |
| ------------------------------------------------------------------------- |
| a. Service account appears in the list |
| b. Verify email: wxcc-ccai-connector@abhavtech-wxcc-ai.iam... |
| |
| [!]️ CHECKPOINT: |
| Expected: Service account visible in list with correct roles |
| |
| Step 7: Record Service Account Information |
| ------------------------------------------------------------------------- |
| Document for WxCC configuration: |
| |
| Service Account Email: ___________________________________ |
| (Full email including @...iam.gserviceaccount.com) |
| |
| [OK] PROCEDURE COMPLETE - Continue to Procedure 3.5.1 |
| |
+-----------------------------------------------------------------------------+
3.5 Generating Service Account Key¶
Procedure: Generate and Download JSON Key¶
+-----------------------------------------------------------------------------+
| PROCEDURE 3.5.1: GENERATE SERVICE ACCOUNT KEY |
+-----------------------------------------------------------------------------+
| |
| Estimated Time: 5 minutes |
| Prerequisites: Procedure 3.3.1 completed |
| |
| [!]️ SECURITY WARNING: |
| ------------------------------------------------------------------------- |
| The JSON key file contains sensitive credentials. Handle with care: |
| * Download only to secure, encrypted storage |
| * Never commit to version control (git) |
| * Never share via email or chat |
| * Store in secure credential management system |
| |
| ======================================================================= |
| |
| Step 1: Navigate to Service Account |
| ------------------------------------------------------------------------- |
| a. Go to: IAM & Admin -> Service Accounts |
| b. Click on "wxcc-ccai-connector" service account |
| |
| Step 2: Navigate to Keys Tab |
| ------------------------------------------------------------------------- |
| a. Click "KEYS" tab (top of page) |
| b. Current keys (if any) are displayed |
| |
| Step 3: Create New Key |
| ------------------------------------------------------------------------- |
| a. Click "ADD KEY" dropdown |
| b. Select "Create new key" |
| c. Key type dialog opens |
| |
| Step 4: Select Key Type |
| ------------------------------------------------------------------------- |
| a. Select: * JSON (recommended) |
| (Do NOT select P12) |
| b. Click "CREATE" |
| |
| Step 5: Download Key File |
| ------------------------------------------------------------------------- |
| a. JSON key file automatically downloads |
| b. Filename format: abhavtech-wxcc-ai-[key-id].json |
| c. Browser shows "Private key saved to your computer" |
| |
| [!]️ This is the ONLY time you can download this key. |
| If lost, you must create a new key. |
| |
| Step 6: Secure the Key File |
| ------------------------------------------------------------------------- |
| a. Move file from Downloads to secure location |
| b. Rename to: abhavtech-wxcc-ccai-key.json |
| c. Recommended storage: |
| - Corporate password manager (e.g., 1Password, Vault) |
| - Encrypted USB drive for offline backup |
| - AWS Secrets Manager / Azure Key Vault / GCP Secret Manager |
| |
| Step 7: Verify Key Contents |
| ------------------------------------------------------------------------- |
| Open JSON file in text editor and verify it contains: |
| |
| { |
| "type": "service_account", |
| "project_id": "abhavtech-wxcc-ai", |
| "private_key_id": "...", |
| "private_key": "-----BEGIN PRIVATE KEY-----\n...", |
| "client_email": "wxcc-ccai-connector@abhavtech-wxcc-ai...", |
| "client_id": "...", |
| "auth_uri": "https://accounts.google.com/o/oauth2/auth", |
| "token_uri": "https://oauth2.googleapis.com/token", |
| ... |
| } |
| |
| [!]️ CHECKPOINT: |
| Expected: JSON file downloaded and contains all required fields |
| If download failed: Repeat from Step 3 |
| |
| Step 8: Record Key Information |
| ------------------------------------------------------------------------- |
| Document (do NOT record the private key itself): |
| |
| Key ID: _______________________________ |
| Created Date: _______________________________ |
| Storage Location: _______________________________ |
| |
| [OK] PROCEDURE COMPLETE - Key ready for WxCC CCAI Connector configuration |
| |
+-----------------------------------------------------------------------------+
4.1 Creating a Dialogflow CX Agent¶
Procedure: Create Dialogflow CX Agent¶
+-----------------------------------------------------------------------------+
| PROCEDURE 4.1.1: CREATE DIALOGFLOW CX AGENT |
+-----------------------------------------------------------------------------+
| |
| Estimated Time: 15 minutes |
| Prerequisites: Section 3 procedures completed |
| |
| Agent Details: |
| ------------------------------------------------------------------------- |
| Display Name: Abhi VA |
| Location: asia-south1 (Mumbai) |
| Default Language: English (en) |
| Time Zone: Asia/Kolkata (IST) |
| |
| ======================================================================= |
| |
| Step 1: Navigate to Dialogflow CX Console |
| ------------------------------------------------------------------------- |
| a. Open browser: https://dialogflow.cloud.google.com/cx |
| b. Sign in with your GCP account |
| c. Select project: "Abhavtech WxCC AI" from dropdown |
| |
| Alternative navigation from GCP Console: |
| a. Navigation Menu (☰) -> Artificial Intelligence -> Dialogflow CX |
| |
| [Screenshot placeholder: Dialogflow CX Console - Agent List] |
| |
| Step 2: Initiate Agent Creation |
| ------------------------------------------------------------------------- |
| a. Click "Create agent" button |
| b. Agent creation dialog opens |
| |
| Step 3: Configure Agent Basics |
| ------------------------------------------------------------------------- |
| a. Enter the following values: |
| |
| Field | Value |
| -------------------+----------------------------------------- |
| Display name | Abhi VA |
| Location | asia-south1 (Mumbai, India) |
| |
| [!]️ IMPORTANT: Location cannot be changed after creation. |
| Select asia-south1 for India data residency compliance. |
| |
| Step 4: Configure Language Settings |
| ------------------------------------------------------------------------- |
| a. Default language: English - en |
| b. Time zone: (GMT+5:30) Asia/Kolkata |
| |
| NOTE: Additional languages (Hindi) will be added after creation |
| |
| Step 5: Create Agent |
| ------------------------------------------------------------------------- |
| a. Click "Create" button |
| b. Wait for agent creation (30-60 seconds) |
| c. Agent opens in Dialogflow CX Console |
| |
| Step 6: Verify Agent Creation |
| ------------------------------------------------------------------------- |
| Verify the following in the Dialogflow CX Console: |
| |
| [ ] Agent name "Abhi VA" appears in header |
| [ ] Location shows "asia-south1" |
| [ ] Default Start Flow is created |
| [ ] Build and Manage sections visible in left panel |
| |
| [!]️ CHECKPOINT: |
| Expected: Agent created in asia-south1 region |
| If wrong region: Delete agent and recreate with correct region |
| |
| Step 7: Record Agent Information |
| ------------------------------------------------------------------------- |
| Document for WxCC configuration: |
| |
| Agent Display Name: _______________________________ |
| Agent ID: _______________________________ |
| Location: _______________________________ |
| |
| To find Agent ID: |
| a. Click Agent Settings (gear icon) |
| b. Agent ID is shown in the settings panel |
| |
| [OK] PROCEDURE COMPLETE - Continue to Procedure 4.2.1 |
| |
+-----------------------------------------------------------------------------+
4.2 Configuring Agent Settings¶
Procedure: Configure Agent Settings¶
+-----------------------------------------------------------------------------+
| PROCEDURE 4.2.1: CONFIGURE DIALOGFLOW CX AGENT SETTINGS |
+-----------------------------------------------------------------------------+
| |
| Estimated Time: 10 minutes |
| |
| ======================================================================= |
| |
| Step 1: Access Agent Settings |
| ------------------------------------------------------------------------- |
| a. In Dialogflow CX Console, with agent open |
| b. Click gear icon (⚙️) next to agent name |
| c. Settings panel opens |
| |
| Step 2: Configure General Settings |
| ------------------------------------------------------------------------- |
| a. Navigate to: General -> General |
| b. Configure: |
| |
| Field | Value |
| -------------------------+---------------------------------- |
| Display name | Abhi VA (no change) |
| Default language | en (English) |
| Time zone | Asia/Kolkata |
| Description | Abhavtech Virtual Agent for |
| | complex interactions |
| |
| Step 3: Add Hindi Language Support |
| ------------------------------------------------------------------------- |
| a. Navigate to: General -> Languages |
| b. Click "+ Add language" |
| c. Select: Hindi (hi) |
| d. Click "Save" |
| |
| Enabled Languages: |
| [ ] English (en) - Default |
| [ ] Hindi (hi) |
| |
| Step 4: Configure Speech Settings |
| ------------------------------------------------------------------------- |
| a. Navigate to: Speech and IVR -> Speech settings |
| b. Configure: |
| |
| Field | Value |
| ------------------------------+------------------------------- |
| Speech model | phone_call (enhanced) |
| Enable speech adaptation | [x] Enabled |
| Audio encoding | AUDIO_ENCODING_LINEAR_16 |
| Sample rate | 8000 Hz (telephony) |
| Enable automatic punctuation | [x] Enabled |
| |
| Step 5: Configure Text-to-Speech Settings |
| ------------------------------------------------------------------------- |
| a. Navigate to: Speech and IVR -> Synthesize speech settings |
| b. Configure for English: |
| |
| Field | Value |
| -------------------+----------------------------------------- |
| Voice name | en-IN-Neural2-A (Male, Indian English) |
| Speaking rate | 1.0 |
| Pitch | 0 |
| |
| c. Configure for Hindi: |
| Click "Add language" -> Hindi |
| |
| Field | Value |
| -------------------+----------------------------------------- |
| Voice name | hi-IN-Neural2-A (Male) |
| Speaking rate | 1.0 |
| Pitch | 0 |
| |
| Step 6: Configure Advanced Speech Settings |
| ------------------------------------------------------------------------- |
| a. Navigate to: Speech and IVR -> Advanced speech settings |
| b. Configure: |
| |
| Field | Value |
| ------------------------------+------------------------------- |
| No-speech timeout | 5 seconds |
| End of speech timeout | 2 seconds |
| Max speech length | 60 seconds |
| Barge-in | [x] Enabled |
| Enable partial responses | [x] Enabled |
| |
| Step 7: Configure Logging |
| ------------------------------------------------------------------------- |
| a. Navigate to: General -> Logging |
| b. Enable: |
| [x] Enable Cloud Logging |
| [x] Enable conversation history |
| [x] Enable interaction logging |
| |
| Step 8: Save All Settings |
| ------------------------------------------------------------------------- |
| a. Click "Save" button |
| b. Wait for confirmation |
| |
| [!]️ CHECKPOINT: |
| Expected: All settings saved, Hindi language added |
| Verify: Two languages shown in Languages section |
| |
| [OK] PROCEDURE COMPLETE - Continue to Procedure 4.3.1 |
| |
+-----------------------------------------------------------------------------+
4.3 Creating Intents¶
Procedure: Create Intent - order.status¶
+-----------------------------------------------------------------------------+
| PROCEDURE 4.3.1: CREATE INTENT - order.status |
+-----------------------------------------------------------------------------+
| |
| Estimated Time: 15 minutes |
| Prerequisites: Procedure 4.2.1 completed |
| |
| ======================================================================= |
| |
| Step 1: Navigate to Intents |
| ------------------------------------------------------------------------- |
| a. In Dialogflow CX Console, with "Abhi VA" agent open |
| b. In left panel under "Build", click "Intents" |
| c. Intents list displays (shows Default Welcome Intent, etc.) |
| |
| Step 2: Create New Intent |
| ------------------------------------------------------------------------- |
| a. Click "+ Create" button |
| b. Intent creation page opens |
| |
| Step 3: Configure Intent Name |
| ------------------------------------------------------------------------- |
| a. Display name: order.status |
| b. Description: Customer checking order status |
| |
| Step 4: Add Training Phrases |
| ------------------------------------------------------------------------- |
| a. In "Training phrases" section |
| b. Click "+ Add training phrase" |
| c. Add each phrase (press Enter after each): |
| |
| English Training Phrases (20): |
| ----------------------------------------------------------- |
| 1. What is my order status |
| 2. Check my order |
| 3. Where is my order |
| 4. Order status |
| 5. Track my order |
| 6. I want to check my order |
| 7. What happened to my order |
| 8. Has my order shipped |
| 9. When will my order arrive |
| 10. I placed an order and want to know the status |
| 11. Check order ORD-12345 |
| 12. Status of order number 12345 |
| 13. Where is order ORD-67890 |
| 14. My order hasn't arrived |
| 15. Order inquiry |
| 16. I need to track a package |
| 17. Did my order ship yet |
| 18. Check on my recent order |
| 19. Order tracking |
| 20. Is my order on the way |
| |
| Step 5: Annotate Entities in Training Phrases |
| ------------------------------------------------------------------------- |
| For phrases containing order numbers: |
| |
| a. In phrase "Check order ORD-12345": |
| - Highlight "ORD-12345" |
| - Click highlighted text |
| - Select entity: @order_number (or create new) |
| |
| b. In phrase "Status of order number 12345": |
| - Highlight "12345" |
| - Assign entity: @order_number |
| |
| c. In phrase "Where is order ORD-67890": |
| - Highlight "ORD-67890" |
| - Assign entity: @order_number |
| |
| Step 6: Add Hindi Training Phrases |
| ------------------------------------------------------------------------- |
| a. Click language dropdown (shows "en") |
| b. Select "hi" (Hindi) |
| c. Add Hindi phrases: |
| |
| Hindi Training Phrases (10): |
| ----------------------------------------------------------- |
| 1. मेरे ऑर्डर का स्टेटस क्या है |
| 2. मेरा ऑर्डर कहाँ है |
| 3. ऑर्डर स्टेटस चेक करो |
| 4. मेरा ऑर्डर ट्रैक करो |
| 5. ऑर्डर कब आएगा |
| 6. मेरा पार्सल कहाँ है |
| 7. ऑर्डर नंबर बारह तीन चार पांच का स्टेटस |
| 8. शिपमेंट का स्टेटस |
| 9. डिलीवरी कब होगी |
| 10. ऑर्डर अभी तक नहीं आया |
| |
| Step 7: Save Intent |
| ------------------------------------------------------------------------- |
| a. Click "Save" button (top of page) |
| b. Wait for "Intent saved" confirmation |
| |
| [!]️ CHECKPOINT: |
| Expected: Intent "order.status" saved with 30 training phrases |
| Verify: Intent appears in Intents list |
| |
| [OK] PROCEDURE COMPLETE - Repeat for remaining intents |
| (See Appendix 10-C for complete training phrase library) |
| |
+-----------------------------------------------------------------------------+
4.7 Configuring Webhooks¶
Procedure: Configure Webhook for Fulfillment¶
+-----------------------------------------------------------------------------+
| PROCEDURE 4.7.1: CONFIGURE DIALOGFLOW CX WEBHOOK |
+-----------------------------------------------------------------------------+
| |
| Estimated Time: 15 minutes |
| Prerequisites: Webhook endpoint deployed (see Appendix 10-B) |
| |
| Webhook Details: |
| ------------------------------------------------------------------------- |
| Name: abhavtech-fulfillment |
| URL: https://api.abhavtech.com/dialogflow/webhook |
| (OR Cloud Function URL if using GCP) |
| |
| ======================================================================= |
| |
| Step 1: Navigate to Webhooks |
| ------------------------------------------------------------------------- |
| a. In Dialogflow CX Console, with agent open |
| b. In left panel under "Manage", click "Webhooks" |
| c. Webhooks list displays |
| |
| Step 2: Create New Webhook |
| ------------------------------------------------------------------------- |
| a. Click "+ Create" button |
| b. Webhook configuration page opens |
| |
| Step 3: Configure Webhook Details |
| ------------------------------------------------------------------------- |
| a. Enter the following: |
| |
| Field | Value |
| -------------------+----------------------------------------- |
| Display name | abhavtech-fulfillment |
| Timeout | 10 seconds |
| Webhook URL | https://api.abhavtech.com/dialogflow/ |
| | webhook |
| |
| Step 4: Configure Authentication (if required) |
| ------------------------------------------------------------------------- |
| a. If webhook requires authentication: |
| - Click "Authentication" section |
| - Select authentication type: |
| |
| Option A - API Key: |
| * API key header |
| Header name: X-API-Key |
| API key: [your-api-key] |
| |
| Option B - OAuth: |
| * OAuth |
| Client ID: [client-id] |
| Client Secret: [client-secret] |
| Token URL: [token-endpoint] |
| |
| Option C - Service Account (for Google Cloud): |
| * Service agent auth header |
| (Uses Dialogflow's service account) |
| |
| Step 5: Configure Subresource (for specific use cases) |
| ------------------------------------------------------------------------- |
| a. For order lookups: |
| - Click "+ Add subresource" |
| - Tag: order_lookup |
| - URL suffix: /orders |
| - Full URL becomes: https://api.abhavtech.com/dialogflow/ |
| webhook/orders |
| |
| b. For account lookups: |
| - Click "+ Add subresource" |
| - Tag: account_lookup |
| - URL suffix: /accounts |
| |
| Step 6: Test Webhook Connection |
| ------------------------------------------------------------------------- |
| a. Click "Test webhook" button (if available) |
| b. Or save and test via simulator |
| |
| Step 7: Save Webhook |
| ------------------------------------------------------------------------- |
| a. Click "Save" button |
| b. Webhook appears in webhooks list |
| |
| Step 8: Link Webhook to Intent Fulfillment |
| ------------------------------------------------------------------------- |
| a. Navigate to the relevant page (e.g., Order Status page) |
| b. In Entry fulfillment section: |
| - Click "Add fulfillment" |
| - Select "Webhook" |
| - Select: abhavtech-fulfillment |
| - Tag: order_lookup |
| c. Save page |
| |
| [!]️ CHECKPOINT: |
| Expected: Webhook configured and linked to relevant pages |
| Test: Use simulator with order number to verify webhook call |
| |
| [OK] PROCEDURE COMPLETE |
| |
+-----------------------------------------------------------------------------+
5.1 Creating CCAI Configuration¶
Procedure: Create CCAI Connector in Control Hub¶
+-----------------------------------------------------------------------------+
| PROCEDURE 5.1.1: CREATE CCAI CONNECTOR FOR DIALOGFLOW CX |
+-----------------------------------------------------------------------------+
| |
| Estimated Time: 15 minutes |
| Prerequisites: |
| - Section 3 completed (GCP project, service account, key) |
| - Section 4 completed (Dialogflow CX agent created) |
| - Service account JSON key file available |
| |
| ======================================================================= |
| |
| Step 1: Navigate to Virtual Agent in Control Hub |
| ------------------------------------------------------------------------- |
| a. Log into Control Hub: https://admin.webex.com |
| b. Navigate: Services -> Contact Center -> Features |
| c. Click "Virtual Agent" |
| |
| Step 2: Create New Virtual Agent (CCAI Type) |
| ------------------------------------------------------------------------- |
| a. Click "+ Create Virtual Agent" |
| b. Select Agent Type: * Dialogflow CX |
| c. Click "Next" |
| |
| Step 3: Configure CCAI Integration |
| ------------------------------------------------------------------------- |
| a. Enter the following values: |
| |
| Field | Value |
| ------------------------+------------------------------------ |
| Name | Abhi_Advanced_VA |
| Description | Dialogflow CX agent for complex |
| | conversations |
| |
| Step 4: Upload Service Account Key |
| ------------------------------------------------------------------------- |
| a. In "Google Cloud Service Account" section |
| b. Click "Upload JSON key file" |
| c. Select your service account key file: |
| abhavtech-wxcc-ccai-key.json |
| d. Wait for validation (system verifies credentials) |
| |
| [!]️ If validation fails: |
| - Verify key file is correct JSON format |
| - Verify service account has Dialogflow API Admin role |
| - Verify project has Dialogflow API enabled |
| |
| Step 5: Select Dialogflow CX Agent |
| ------------------------------------------------------------------------- |
| After key validation succeeds: |
| |
| a. "Project" dropdown populates automatically |
| - Verify: abhavtech-wxcc-ai |
| |
| b. Click "Location" dropdown |
| - Select: asia-south1 |
| |
| c. Click "Agent" dropdown |
| - Select: Abhi VA |
| |
| d. "Environment" dropdown appears |
| - Select: draft (for testing) OR production (for live) |
| - For initial setup, select: draft |
| |
| Field | Value |
| -------------------+----------------------------------------- |
| Project | abhavtech-wxcc-ai |
| Location | asia-south1 |
| Agent | Abhi VA |
| Environment | draft |
| |
| Step 6: Configure Voice Settings |
| ------------------------------------------------------------------------- |
| a. Click "Next" to proceed to Voice Settings |
| b. Configure: |
| |
| Field | Value |
| -------------------------+---------------------------------- |
| Input language | en-IN (English India) |
| Output voice | en-IN-Neural2-A (Male) |
| Speech model | phone_call |
| Single utterance mode | O Disabled (allow multi-turn) |
| End of speech timeout | 2000 ms |
| No speech timeout | 5000 ms |
| Barge-in | * Enabled |
| |
| Step 7: Add Additional Language (Hindi) |
| ------------------------------------------------------------------------- |
| a. Click "+ Add language" |
| b. Configure: |
| |
| Field | Value |
| -------------------------+---------------------------------- |
| Input language | hi-IN (Hindi India) |
| Output voice | hi-IN-Neural2-A (Male) |
| |
| Step 8: Review and Create |
| ------------------------------------------------------------------------- |
| a. Click "Next" to review configuration |
| b. Verify all settings are correct |
| c. Click "Create" |
| d. Wait for creation (30-60 seconds) |
| |
| Step 9: Verify Creation |
| ------------------------------------------------------------------------- |
| a. "Abhi_Advanced_VA" appears in Virtual Agent list |
| b. Status shows "Active" (green indicator) |
| c. Type shows "Dialogflow CX" |
| |
| [!]️ CHECKPOINT: |
| Expected: CCAI connector created with "Active" status |
| If "Error" status: Check GCP credentials and API enablement |
| |
| Step 10: Record Configuration |
| ------------------------------------------------------------------------- |
| Document for flow configuration: |
| |
| Virtual Agent Name: _______________________________ |
| GCP Project: _______________________________ |
| Dialogflow Agent: _______________________________ |
| Region: _______________________________ |
| Environment: _______________________________ |
| |
| [OK] PROCEDURE COMPLETE - Continue to Section 6 for flow integration |
| |
+-----------------------------------------------------------------------------+
6.1 Adding Virtual Agent V2 Node¶
Procedure: Add Virtual Agent Node to Flow¶
+-----------------------------------------------------------------------------+
| PROCEDURE 6.1.1: ADD VIRTUAL AGENT V2 NODE TO FLOW |
+-----------------------------------------------------------------------------+
| |
| Estimated Time: 20 minutes |
| Prerequisites: |
| - Section 2 completed (Webex AI Agent created) |
| - Section 5 completed (CCAI Connector created) |
| - Baseline flow exists (India_MainMenu_Flow_v1) |
| |
| ======================================================================= |
| |
| Step 1: Access Flow Designer |
| ------------------------------------------------------------------------- |
| a. In Control Hub: Services -> Contact Center -> Flows |
| b. Find flow: India_MainMenu_Flow_v1 |
| c. Click flow name to open |
| d. Click "Edit" button to enter edit mode |
| |
| [Screenshot placeholder: Flow Designer canvas] |
| |
| Step 2: Locate Insertion Point |
| ------------------------------------------------------------------------- |
| a. Identify where VA node should be inserted |
| b. For India flow: After initial greeting, before DTMF menu |
| |
| Current Flow: |
| [Start] -> [Play Welcome] -> [DTMF Menu] -> [Queue] |
| |
| Target Flow: |
| [Start] -> [Play Welcome] -> [Virtual Agent] -> [DTMF Menu] -> [Queue]|
| ^ INSERT HERE |
| |
| Step 3: Add Virtual Agent V2 Node |
| ------------------------------------------------------------------------- |
| a. In left panel, expand "Activities" |
| b. Locate "Virtual Agent V2" activity |
| (Under "Virtual Agent" category) |
| c. Drag "Virtual Agent V2" onto canvas |
| d. Position between Play Welcome and DTMF Menu |
| |
| Step 4: Connect Node to Flow |
| ------------------------------------------------------------------------- |
| a. Delete existing connection from Play Welcome to DTMF Menu |
| - Click the connection line |
| - Press Delete key |
| |
| b. Connect Play Welcome -> Virtual Agent V2 |
| - Click output port of Play Welcome |
| - Drag to input port of Virtual Agent V2 |
| |
| c. The Virtual Agent V2 node has multiple output ports: |
| - Handled (self-service complete) |
| - Escalated (transfer to agent) |
| - Error (technical failure) |
| |
| Step 5: Configure Virtual Agent V2 Node |
| ------------------------------------------------------------------------- |
| a. Click on Virtual Agent V2 node to select |
| b. Properties panel opens on right |
| |
| Step 6: Configure Contact Center AI Config |
| ------------------------------------------------------------------------- |
| a. In "Contact Center AI Config" dropdown: |
| - Select: Abhi_Advanced_VA (Dialogflow CX) |
| |
| b. For EMEA/Americas flows, you would select: |
| - Abhi_Simple_VA (Webex AI Agent) |
| |
| Flow | Virtual Agent Config |
| ------------------------+------------------------------------ |
| India_MainMenu_Flow_v1 | Abhi_Advanced_VA (Dialogflow CX) |
| EMEA_MainMenu_Flow_v1 | Abhi_Simple_VA (Webex AI Agent) |
| Americas_MainMenu_Flow | Abhi_Simple_VA (Webex AI Agent) |
| |
| Step 7: Configure Input Variables |
| ------------------------------------------------------------------------- |
| a. Expand "Input Variables" section |
| b. Add variables to pass to Virtual Agent: |
| |
| Click "+ Add Variable Mapping" |
| |
| VA Variable | Flow Variable |
| -----------------+----------------------------------------- |
| customer_ani | {{NewPhoneContact.ANI}} |
| customer_name | {{Global_CustomerName}} (if available) |
| session_id | {{NewPhoneContact.InteractionId}} |
| language | {{Customer_Language}} (if set) |
| |
| Step 8: Configure Conversation Settings |
| ------------------------------------------------------------------------- |
| a. Expand "Settings" section |
| b. Configure: |
| |
| Field | Value |
| ---------------------------+--------------------------------- |
| Max Conversation Turns | 10 |
| No-Input Timeout (ms) | 5000 |
| Speech Complete Timeout | 2000 |
| Enable Barge-In | [x] Enabled |
| Enable DTMF Collection | [x] Enabled |
| |
| Step 9: Configure Language Override (Optional) |
| ------------------------------------------------------------------------- |
| a. If language was detected earlier in flow: |
| - Language Override: {{Customer_Language}} |
| b. Otherwise leave empty for auto-detection |
| |
| Step 10: Save Node Configuration |
| ------------------------------------------------------------------------- |
| a. Click outside the node to deselect |
| b. Configuration is auto-saved |
| |
| [!]️ CHECKPOINT: |
| Expected: Virtual Agent V2 node configured with Dialogflow CX |
| Verify: Node shows "Abhi_Advanced_VA" in configuration |
| |
| [OK] PROCEDURE COMPLETE - Continue to Procedure 6.3.1 for exit handling |
| |
+-----------------------------------------------------------------------------+
6.3 Handling Exit Events¶
Procedure: Configure Exit Event Handling¶
+-----------------------------------------------------------------------------+
| PROCEDURE 6.3.1: CONFIGURE VIRTUAL AGENT EXIT HANDLING |
+-----------------------------------------------------------------------------+
| |
| Estimated Time: 15 minutes |
| |
| ======================================================================= |
| |
| Step 1: Identify Exit Ports |
| ------------------------------------------------------------------------- |
| The Virtual Agent V2 node has three exit ports: |
| |
| | Port | Triggers When | Action |
| |------------|------------------------------------|--------------|
| | Handled | VA successfully resolved query | End call or |
| | | (containment) | survey |
| |------------|------------------------------------|--------------|
| | Escalated | VA transfers to human agent | Queue to |
| | | (intent or request) | agent |
| |------------|------------------------------------|--------------|
| | Error | Technical failure | Fallback |
| | | (timeout, API error) | to DTMF |
| |
| Step 2: Configure "Handled" Exit Path |
| ------------------------------------------------------------------------- |
| For successfully contained interactions: |
| |
| a. Connect "Handled" port -> Survey or Disconnect |
| - Drag from "Handled" port |
| - Connect to: Survey_PostCall_v1 (subflow) |
| OR Disconnect node |
| |
| b. Optionally add Set Variable before disconnect: |
| - VA_Outcome = "contained" |
| - VA_LastIntent = {{VirtualAgent.LastIntent}} |
| |
| Flow for Handled: |
| [Virtual Agent] ->(Handled)-> [Set Variables] -> [Survey] -> [End] |
| |
| Step 3: Configure "Escalated" Exit Path |
| ------------------------------------------------------------------------- |
| For agent transfer requests: |
| |
| a. Add "Set Variable" node after Escalated port |
| - Drag Set Variable activity to canvas |
| - Connect "Escalated" -> Set Variable |
| |
| b. Configure variables for agent context: |
| |
| Variable Name | Value |
| ---------------------+------------------------------------- |
| VA_Escalation_Reason | {{VirtualAgent.ExitReason}} |
| VA_Last_Intent | {{VirtualAgent.LastIntent}} |
| VA_Sentiment | {{VirtualAgent.Sentiment}} |
| VA_Confidence | {{VirtualAgent.IntentConfidence}} |
| VA_Transcript_URL | {{VirtualAgent.TranscriptURL}} |
| VA_Collected_Data | {{VirtualAgent.CustomPayload}} |
| |
| c. Add routing logic based on intent/sentiment: |
| |
| Add "Condition" node: |
| - IF VA_Sentiment == "negative" AND score < -0.5 |
| -> Route to: Sentiment_Priority_Queue |
| - ELSE IF VA_Last_Intent == "billing.inquiry" |
| -> Route to: Billing_Queue |
| - ELSE |
| -> Route to: VA_Escalation_Queue |
| |
| Flow for Escalated: |
| [Virtual Agent] ->(Escalated)-> [Set Vars] -> [Condition] -> [Queue] |
| |
| Step 4: Configure "Error" Exit Path |
| ------------------------------------------------------------------------- |
| For technical failures: |
| |
| a. Add "Play Message" node for error announcement |
| - Connect "Error" -> Play Message |
| - Message: "We're experiencing technical difficulties. |
| Please use our menu to direct your call." |
| |
| b. Connect Play Message -> DTMF Menu (fallback) |
| |
| c. Add logging for troubleshooting: |
| - Add Set Variable before DTMF Menu |
| - Error_Code = {{VirtualAgent.ErrorMessage}} |
| |
| Flow for Error: |
| [Virtual Agent] ->(Error)-> [Log Error] -> [Play Msg] -> [DTMF Menu] |
| |
| Step 5: Verify All Paths Connected |
| ------------------------------------------------------------------------- |
| Check that all three exit ports have connections: |
| |
| [ ] Handled -> Survey/Disconnect |
| [ ] Escalated -> Queue (via routing logic) |
| [ ] Error -> DTMF Fallback |
| |
| Step 6: Save and Validate Flow |
| ------------------------------------------------------------------------- |
| a. Click "Validate" button |
| b. Review any warnings or errors |
| c. Fix any disconnected paths |
| d. Click "Save" when validation passes |
| |
| [!]️ CHECKPOINT: |
| Expected: Flow validates without errors |
| All VA exit paths connected to appropriate destinations |
| |
| [OK] PROCEDURE COMPLETE - Flow ready for testing |
| |
+-----------------------------------------------------------------------------+
7.1 Enabling Agent Assist¶
Procedure: Enable Cisco AI Assistant for Agents¶
+-----------------------------------------------------------------------------+
| PROCEDURE 7.1.1: ENABLE CISCO AI ASSISTANT (AGENT ASSIST) |
+-----------------------------------------------------------------------------+
| |
| Estimated Time: 15 minutes |
| Prerequisites: WxCC tenant with Premium Agent licenses |
| |
| ======================================================================= |
| |
| Step 1: Navigate to Agent Assist Settings |
| ------------------------------------------------------------------------- |
| a. Log into Control Hub: https://admin.webex.com |
| b. Navigate: Services -> Contact Center -> Features |
| c. Click "AI Assistant" (or "Agent Assist") |
| |
| [Screenshot placeholder: Control Hub - AI Assistant] |
| |
| Step 2: Enable AI Assistant |
| ------------------------------------------------------------------------- |
| a. Toggle "Enable AI Assistant" to ON |
| b. Confirmation dialog appears |
| c. Click "Enable" |
| |
| Step 3: Configure General Settings |
| ------------------------------------------------------------------------- |
| a. In Settings section, configure: |
| |
| Field | Value |
| ---------------------------+--------------------------------- |
| Enable for Voice | [x] Enabled |
| Enable for Digital | [x] Enabled |
| Default Language | English (en-US) |
| Additional Languages | [x] Hindi (hi-IN) |
| |
| Step 4: Configure Feature Toggles |
| ------------------------------------------------------------------------- |
| Enable the following features: |
| |
| Feature | Status | Description |
| ---------------------------+---------+-------------------------|
| Real-time Transcription | [x] On | Live call transcript |
| Conversation Summary | [x] On | AI-generated summary |
| Suggested Responses | [x] On | Agent response hints |
| Sentiment Analysis | [x] On | Customer mood tracking |
| Knowledge Suggestions | [x] On | KB article suggestions |
| Auto Wrap-up Codes | [x] On | AI disposition suggest |
| Next Best Action | O Off | (Phase 3 - needs data) |
| |
| Step 5: Configure User Profile Assignment |
| ------------------------------------------------------------------------- |
| a. Navigate to: User Profiles section |
| b. Select profiles to enable AI Assistant: |
| |
| Profile | AI Assistant |
| ---------------------------+--------------------------------- |
| Standard_Agent | O Disabled (Voice only agents) |
| Premium_Agent | * Enabled |
| Supervisor | * Enabled |
| |
| c. Click "Save" |
| |
| Step 6: Verify Enablement |
| ------------------------------------------------------------------------- |
| a. Status should show "Active" |
| b. Enabled profiles count should show: 85 |
| (75 Premium + 10 Supervisor) |
| |
| [!]️ CHECKPOINT: |
| Expected: AI Assistant active for Premium and Supervisor profiles |
| |
| [OK] PROCEDURE COMPLETE - Continue to Knowledge Base setup |
| |
+-----------------------------------------------------------------------------+
7.2 Configuring Knowledge Base¶
Procedure: Create and Populate Knowledge Base¶
+-----------------------------------------------------------------------------+
| PROCEDURE 7.2.1: CREATE KNOWLEDGE BASE FOR AGENT ASSIST |
+-----------------------------------------------------------------------------+
| |
| Estimated Time: 30-60 minutes (depending on content volume) |
| Prerequisites: Procedure 7.1.1 completed |
| |
| Knowledge Base Specification: |
| ------------------------------------------------------------------------- |
| Target Articles: 125 total |
| Categories: Product FAQs, Troubleshooting, Policies, Billing, Company |
| |
| ======================================================================= |
| |
| Step 1: Navigate to Knowledge Base |
| ------------------------------------------------------------------------- |
| a. In Control Hub: Services -> Contact Center -> Features |
| b. Click "Knowledge Base" (under AI Assistant) |
| c. Knowledge Base management page opens |
| |
| Step 2: Create Knowledge Base |
| ------------------------------------------------------------------------- |
| a. Click "+ Create Knowledge Base" |
| b. Enter: |
| |
| Field | Value |
| -------------------+----------------------------------------- |
| Name | Abhavtech_Support_KB |
| Description | Knowledge base for agent assistance |
| Default Language | English |
| |
| c. Click "Create" |
| |
| Step 3: Create Categories |
| ------------------------------------------------------------------------- |
| a. Click "Categories" tab |
| b. Click "+ Add Category" for each: |
| |
| Category Name | Target Articles |
| --------------------------+---------------------------------- |
| Product_FAQs | 50 |
| Troubleshooting_Guides | 30 |
| Policies_Procedures | 20 |
| Billing_FAQs | 15 |
| Company_Information | 10 |
| |
| Step 4: Add Articles - Manual Entry |
| ------------------------------------------------------------------------- |
| For each article: |
| |
| a. Click "Articles" tab |
| b. Click "+ Add Article" |
| c. Fill in article details: |
| |
| Field | Example Value |
| -------------------+----------------------------------------- |
| Title | How to Reset Product A Password |
| Category | Product_FAQs |
| Content | [Article body text] |
| Keywords | password, reset, ProductA, login |
| Language | English |
| |
| d. Click "Save" |
| e. Repeat for all articles |
| |
| Step 5: Bulk Import (Alternative Method) |
| ------------------------------------------------------------------------- |
| For large volume import: |
| |
| a. Click "Import" button |
| b. Download CSV template |
| c. Populate CSV with articles: |
| |
| CSV Format: |
| title,category,content,keywords,language |
| "How to Reset...","Product_FAQs","Step 1...","pwd,reset","en" |
| |
| d. Upload completed CSV |
| e. Review import results |
| f. Fix any errors and re-import if needed |
| |
| Step 6: Add Hindi Language Articles |
| ------------------------------------------------------------------------- |
| a. For high-priority articles, add Hindi versions: |
| b. Click article -> "Add Translation" |
| c. Select language: Hindi (hi) |
| d. Enter translated title and content |
| e. Save translation |
| |
| Priority Hindi Translations: |
| [ ] Top 20 Product FAQs |
| [ ] Top 10 Troubleshooting Guides |
| [ ] All Billing FAQs (15) |
| |
| Step 7: Publish Knowledge Base |
| ------------------------------------------------------------------------- |
| a. Review all articles for accuracy |
| b. Click "Publish" button |
| c. Select "Publish All" |
| d. Confirm publication |
| |
| Step 8: Link to AI Assistant |
| ------------------------------------------------------------------------- |
| a. Navigate back to AI Assistant settings |
| b. In "Knowledge Base" section |
| c. Select: Abhavtech_Support_KB |
| d. Click "Save" |
| |
| [!]️ CHECKPOINT: |
| Expected: Knowledge Base with 125+ articles linked to AI Assistant |
| Verify: Test search returns relevant articles |
| |
| [OK] PROCEDURE COMPLETE - Continue to Suggestions setup |
| |
+-----------------------------------------------------------------------------+
7.3 Setting Up Suggestions¶
Procedure: Configure Agent Response Suggestions¶
+-----------------------------------------------------------------------------+
| PROCEDURE 7.3.1: CONFIGURE SUGGESTED RESPONSES |
+-----------------------------------------------------------------------------+
| |
| Estimated Time: 20 minutes |
| Prerequisites: Procedures 7.1.1 and 7.2.1 completed |
| |
| ======================================================================= |
| |
| Step 1: Navigate to Suggestions Settings |
| ------------------------------------------------------------------------- |
| a. In AI Assistant settings |
| b. Click "Suggestions" section |
| |
| Step 2: Configure Suggestion Sources |
| ------------------------------------------------------------------------- |
| Enable suggestion sources: |
| |
| Source | Status | Priority |
| ---------------------------+---------+------------------------|
| Knowledge Base Articles | [x] On | 1 (Highest) |
| Canned Responses | [x] On | 2 |
| Historical Responses | O Off | (Phase 3) |
| Smart Compose | [x] On | 3 |
| |
| Step 3: Configure Display Settings |
| ------------------------------------------------------------------------- |
| a. Set display parameters: |
| |
| Field | Value |
| ---------------------------+--------------------------------- |
| Max suggestions displayed | 3 |
| Confidence threshold | 0.7 (70%) |
| Auto-refresh interval | 5 seconds |
| Show confidence scores | [x] Yes (for supervisors) |
| |
| Step 4: Create Canned Responses |
| ------------------------------------------------------------------------- |
| a. Click "Canned Responses" sub-section |
| b. Click "+ Add Response" |
| c. Create responses for common scenarios: |
| |
| Response 1: Greeting |
| +-------------------------------------------------------------+ |
| | Name: greeting_standard | |
| | Trigger: call_start | |
| | Text: "Thank you for contacting Abhavtech support. | |
| | My name is [Agent Name]. How may I assist you | |
| | today?" | |
| | Language: English | |
| +-------------------------------------------------------------+ |
| |
| Response 2: Hold Request |
| +-------------------------------------------------------------+ |
| | Name: hold_request | |
| | Trigger: manual | |
| | Text: "I need to look into this for you. Would you | |
| | mind holding for a moment while I check?" | |
| | Language: English | |
| +-------------------------------------------------------------+ |
| |
| Response 3: Escalation |
| +-------------------------------------------------------------+ |
| | Name: escalation_transfer | |
| | Trigger: manual | |
| | Text: "I understand this is a complex issue. Let me | |
| | connect you with a specialist who can better | |
| | assist you." | |
| | Language: English | |
| +-------------------------------------------------------------+ |
| |
| Response 4: Closing |
| +-------------------------------------------------------------+ |
| | Name: closing_standard | |
| | Trigger: call_end | |
| | Text: "Is there anything else I can help you with | |
| | today? Thank you for choosing Abhavtech." | |
| | Language: English | |
| +-------------------------------------------------------------+ |
| |
| d. Add Hindi versions for each response |
| |
| Step 5: Configure Trigger Rules |
| ------------------------------------------------------------------------- |
| a. Click "Trigger Rules" sub-section |
| b. Create rules for automatic suggestions: |
| |
| Rule | Trigger Condition |
| ---------------------------+--------------------------------- |
| Order_Status_Detected | Intent = "order.status" |
| | -> Suggest: Order lookup steps |
| Billing_Issue_Detected | Intent = "billing.inquiry" |
| | -> Suggest: Billing KB articles |
| Negative_Sentiment | Sentiment < -0.5 |
| | -> Suggest: Empathy responses |
| Long_Hold_Time | Hold duration > 60s |
| | -> Suggest: Apology response |
| |
| Step 6: Save Configuration |
| ------------------------------------------------------------------------- |
| a. Click "Save" button |
| b. Verify status shows "Configuration saved" |
| |
| [!]️ CHECKPOINT: |
| Expected: Suggestions configured with KB, canned responses, triggers |
| Test: Verify suggestions appear in Agent Desktop |
| |
| [OK] PROCEDURE COMPLETE - Continue to Conversation Summaries |
| |
+-----------------------------------------------------------------------------+
7.4 Configuring Conversation Summaries¶
Procedure: Configure AI-Generated Summaries¶
+-----------------------------------------------------------------------------+
| PROCEDURE 7.4.1: CONFIGURE CONVERSATION SUMMARIES |
+-----------------------------------------------------------------------------+
| |
| Estimated Time: 15 minutes |
| Prerequisites: Procedure 7.1.1 completed |
| |
| ======================================================================= |
| |
| Step 1: Navigate to Summary Settings |
| ------------------------------------------------------------------------- |
| a. In AI Assistant settings |
| b. Click "Conversation Summaries" section |
| |
| Step 2: Enable Summary Types |
| ------------------------------------------------------------------------- |
| Configure which summaries to generate: |
| |
| Summary Type | Status | Use Case |
| ---------------------------+---------+------------------------|
| VA Handoff Summary | [x] On | VA -> Agent transfer |
| Mid-Call Summary | [x] On | Agent -> Agent transfer |
| Wrap-up Summary | [x] On | Post-call notes |
| Supervisor Summary | [x] On | Call review |
| |
| Step 3: Configure VA Handoff Summary |
| ------------------------------------------------------------------------- |
| This summary appears when Virtual Agent transfers to human: |
| |
| a. Click "VA Handoff Summary" settings |
| b. Configure: |
| |
| Field | Value |
| ---------------------------+--------------------------------- |
| Include in summary: |
| +- Customer intent | [x] Yes |
| +- Collected information | [x] Yes |
| +- VA conversation turns | [x] Yes |
| +- Escalation reason | [x] Yes |
| +- Customer sentiment | [x] Yes |
| +- Recommended actions | [x] Yes |
| |
| Display location | Screen pop + Side panel |
| Auto-expand on transfer | [x] Yes |
| |
| Step 4: Configure Wrap-up Summary |
| ------------------------------------------------------------------------- |
| This summary auto-generates when call ends: |
| |
| a. Click "Wrap-up Summary" settings |
| b. Configure: |
| |
| Field | Value |
| ---------------------------+--------------------------------- |
| Auto-generate | [x] Yes |
| Include: |
| +- Call reason | [x] Yes |
| +- Actions taken | [x] Yes |
| +- Resolution status | [x] Yes |
| +- Follow-up required | [x] Yes |
| +- Customer sentiment | [x] Yes |
| |
| Allow agent edit | [x] Yes |
| Max summary length | 500 characters |
| |
| Step 5: Configure Auto Wrap-up Codes |
| ------------------------------------------------------------------------- |
| AI suggests disposition codes based on conversation: |
| |
| a. Click "Auto Wrap-up" settings |
| b. Enable: [x] Suggest wrap-up codes |
| c. Map intents to wrap-up codes: |
| |
| Detected Intent | Suggested Wrap-up Code |
| ---------------------------+--------------------------------- |
| order.status | Order Inquiry |
| order.track | Shipping/Tracking |
| billing.inquiry | Billing Question |
| support.troubleshoot | Technical Support |
| account.info | Account Management |
| product.inquiry | Product Information |
| (unresolved) | Escalated/Unresolved |
| |
| d. Set confidence threshold: 0.8 (80%) |
| e. Allow agent override: [x] Yes |
| |
| Step 6: Test Summary Generation |
| ------------------------------------------------------------------------- |
| a. Make a test call through VA that escalates to agent |
| b. Accept call as test agent |
| c. Verify VA Handoff Summary appears |
| d. Complete call and verify Wrap-up Summary generates |
| |
| [!]️ CHECKPOINT: |
| Expected: Summaries auto-generate for VA handoffs and wrap-ups |
| Test: Verify summary accuracy matches conversation |
| |
| [OK] PROCEDURE COMPLETE - Section 7 complete |
| |
+-----------------------------------------------------------------------------+
8.1 Component Validation Checklist¶
+-----------------------------------------------------------------------------+
| CHECKLIST 8.1: COMPONENT VALIDATION |
+-----------------------------------------------------------------------------+
| |
| Complete this checklist before proceeding to end-to-end testing. |
| |
| WEBEX AI AGENT (Abhi_Simple_VA): |
| ------------------------------------------------------------------------- |
| [ ] Agent created in Control Hub |
| [ ] Status shows "Published" |
| [ ] 5 intents configured: |
| [ ] greeting.hello (15 training phrases) |
| [ ] hours.location (20 training phrases) |
| [ ] callback.request (18 training phrases) |
| [ ] agent.handoff (25 training phrases) |
| [ ] fallback.default (system) |
| [ ] Simulator tests pass for all intents |
| |
| GCP PROJECT (abhavtech-wxcc-ai): |
| ------------------------------------------------------------------------- |
| [ ] Project created in asia-south1 |
| [ ] Billing account linked |
| [ ] APIs enabled: |
| [ ] Dialogflow API |
| [ ] Cloud Speech-to-Text API |
| [ ] Cloud Text-to-Speech API |
| [ ] Cloud Functions API |
| [ ] Cloud Build API |
| [ ] Service account created: wxcc-ccai-connector |
| [ ] Service account has Dialogflow API Admin role |
| [ ] JSON key file generated and secured |
| |
| DIALOGFLOW CX AGENT (Abhi VA): |
| ------------------------------------------------------------------------- |
| [ ] Agent created in asia-south1 |
| [ ] Languages configured: English (en), Hindi (hi) |
| [ ] Voice settings configured for both languages |
| [ ] Intents created (minimum for testing): |
| [ ] order.status |
| [ ] agent.handoff |
| [ ] Default Welcome Intent |
| [ ] Default Fallback Intent |
| [ ] Webhook configured (if using fulfillment) |
| [ ] Simulator tests pass |
| |
| WXCC CCAI CONNECTOR (Abhi_Advanced_VA): |
| ------------------------------------------------------------------------- |
| [ ] Connector created in Control Hub |
| [ ] Status shows "Active" |
| [ ] Linked to correct GCP project |
| [ ] Linked to correct Dialogflow CX agent |
| [ ] Voice settings configured |
| [ ] Languages configured (en-IN, hi-IN) |
| |
| FLOW DESIGNER INTEGRATION: |
| ------------------------------------------------------------------------- |
| [ ] Virtual Agent V2 node added to flow |
| [ ] Node configured with correct CCAI config |
| [ ] Input variables mapped |
| [ ] All exit paths connected: |
| [ ] Handled -> Survey/Disconnect |
| [ ] Escalated -> Queue |
| [ ] Error -> DTMF Fallback |
| [ ] Flow validates without errors |
| [ ] Flow published |
| |
| All items checked? [ ] YES -> Proceed to End-to-End Testing |
| [ ] NO -> Complete missing items first |
| |
+-----------------------------------------------------------------------------+
8.2 End-to-End Test Procedure¶
+-----------------------------------------------------------------------------+
| PROCEDURE 8.2.1: END-TO-END AI INTEGRATION TEST |
+-----------------------------------------------------------------------------+
| |
| Estimated Time: 30 minutes |
| Test Environment: UAT tenant with test DIDs |
| |
| ======================================================================= |
| |
| TEST 1: DIALOGFLOW CX - ORDER STATUS (INDIA FLOW) |
| ------------------------------------------------------------------------- |
| |
| Step 1: Dial India test number |
| Number: [test DID] |
| |
| Step 2: Listen for welcome prompt |
| Expected: "Hello! Welcome to Abhavtech..." |
| [ ] PASS [ ] FAIL |
| |
| Step 3: Say "Check my order status" |
| Expected: VA asks for order number |
| [ ] PASS [ ] FAIL |
| |
| Step 4: Say "ORD-12345" (test order) |
| Expected: VA provides order status from webhook |
| [ ] PASS [ ] FAIL |
| |
| Step 5: Say "Thank you, goodbye" |
| Expected: VA ends conversation (Handled exit) |
| [ ] PASS [ ] FAIL |
| |
| ------------------------------------------------------------------------- |
| |
| TEST 2: DIALOGFLOW CX - AGENT ESCALATION (INDIA FLOW) |
| ------------------------------------------------------------------------- |
| |
| Step 1: Dial India test number |
| |
| Step 2: After welcome, say "I want to speak to an agent" |
| Expected: VA acknowledges and transfers |
| [ ] PASS [ ] FAIL |
| |
| Step 3: Verify routing |
| Expected: Call routes to VA_Escalation_Queue |
| [ ] PASS [ ] FAIL |
| |
| Step 4: Agent accepts call (use test agent) |
| Expected: Agent sees VA context in screen pop |
| [ ] PASS [ ] FAIL |
| |
| ------------------------------------------------------------------------- |
| |
| TEST 3: WEBEX AI AGENT - HOURS INQUIRY (EMEA FLOW) |
| ------------------------------------------------------------------------- |
| |
| Step 1: Dial EMEA test number |
| |
| Step 2: After welcome, say "What are your business hours" |
| Expected: VA provides hours information |
| [ ] PASS [ ] FAIL |
| |
| Step 3: Say "Thank you" |
| Expected: VA offers additional help or ends |
| [ ] PASS [ ] FAIL |
| |
| ------------------------------------------------------------------------- |
| |
| TEST 4: WEBEX AI AGENT - CALLBACK REQUEST (AMERICAS FLOW) |
| ------------------------------------------------------------------------- |
| |
| Step 1: Dial Americas test number |
| |
| Step 2: After welcome, say "I want a callback" |
| Expected: VA confirms callback number |
| [ ] PASS [ ] FAIL |
| |
| Step 3: Confirm "Yes" |
| Expected: Callback scheduled, call ends |
| [ ] PASS [ ] FAIL |
| |
| Step 4: Verify callback in queue |
| Expected: Callback appears in queue |
| [ ] PASS [ ] FAIL |
| |
| ------------------------------------------------------------------------- |
| |
| TEST 5: FALLBACK HANDLING |
| ------------------------------------------------------------------------- |
| |
| Step 1: Dial any test number |
| |
| Step 2: Say gibberish or unrelated topic |
| Expected: VA asks for clarification |
| [ ] PASS [ ] FAIL |
| |
| Step 3: Repeat gibberish two more times |
| Expected: DTMF menu offered |
| [ ] PASS [ ] FAIL |
| |
| Step 4: Press "0" for agent |
| Expected: Routed to support queue |
| [ ] PASS [ ] FAIL |
| |
| ------------------------------------------------------------------------- |
| |
| TEST SUMMARY: |
| ======================================================================= |
| |
| | Test | Result | Notes | |
| |------------------------------|--------|---------------------------| |
| | Test 1: Order Status | | | |
| | Test 2: Agent Escalation | | | |
| | Test 3: Hours Inquiry | | | |
| | Test 4: Callback Request | | | |
| | Test 5: Fallback Handling | | | |
| |
| Overall Result: [ ] PASS (all tests pass) [ ] FAIL (see notes) |
| |
| Sign-off: |
| Tested by: _______________________ Date: _______________ |
| Reviewed by: _____________________ Date: _______________ |
| |
+-----------------------------------------------------------------------------+
8.3 Troubleshooting Common Issues¶
+-----------------------------------------------------------------------------+
| TROUBLESHOOTING GUIDE: COMMON AI INTEGRATION ISSUES |
+-----------------------------------------------------------------------------+
| |
| ISSUE 1: VIRTUAL AGENT NOT RESPONDING |
| ------------------------------------------------------------------------- |
| Symptoms: |
| * Silence after welcome prompt |
| * Call drops or goes directly to queue |
| |
| Troubleshooting Steps: |
| 1. Check Virtual Agent status in Control Hub |
| -> Must show "Active" or "Published" |
| 2. Verify flow is using correct VA configuration |
| -> Check Virtual Agent V2 node settings |
| 3. Check GCP project APIs are enabled |
| -> Especially Dialogflow API |
| 4. Verify service account key is valid |
| -> Try regenerating key if expired |
| 5. Check network connectivity to Google APIs |
| -> Firewall may be blocking |
| |
| ------------------------------------------------------------------------- |
| |
| ISSUE 2: INTENTS NOT RECOGNIZED |
| ------------------------------------------------------------------------- |
| Symptoms: |
| * VA always triggers fallback |
| * Wrong intent detected |
| |
| Troubleshooting Steps: |
| 1. Test intent in Dialogflow CX simulator |
| -> Verify intent matches in simulator |
| 2. Check training phrases |
| -> Add more variations |
| 3. Review confidence threshold |
| -> May be set too high (try 0.5) |
| 4. Check for conflicting intents |
| -> Similar training phrases in multiple intents |
| 5. Verify language settings |
| -> VA language must match caller language |
| |
| ------------------------------------------------------------------------- |
| |
| ISSUE 3: WEBHOOK TIMEOUT OR FAILURE |
| ------------------------------------------------------------------------- |
| Symptoms: |
| * VA says "having trouble" or similar |
| * Order/account lookups fail |
| |
| Troubleshooting Steps: |
| 1. Check webhook endpoint is accessible |
| -> Test URL directly with curl |
| 2. Verify webhook timeout setting |
| -> Increase if backend is slow (max 30s) |
| 3. Check webhook authentication |
| -> API key or OAuth may be expired |
| 4. Review Cloud Function logs (if GCP) |
| -> GCP Console -> Cloud Functions -> Logs |
| 5. Test webhook with Dialogflow simulator |
| -> Shows actual request/response |
| |
| ------------------------------------------------------------------------- |
| |
| ISSUE 4: CCAI CONNECTOR "ERROR" STATUS |
| ------------------------------------------------------------------------- |
| Symptoms: |
| * Connector shows error in Control Hub |
| * Cannot select in Flow Designer |
| |
| Troubleshooting Steps: |
| 1. Verify service account permissions |
| -> Must have Dialogflow API Admin |
| 2. Check service account key validity |
| -> Keys can be disabled or deleted in GCP |
| 3. Verify Dialogflow CX agent exists |
| -> Agent may have been deleted |
| 4. Check region match |
| -> Connector and agent must be same region |
| 5. Regenerate service account key |
| -> Create new key and re-upload |
| |
| ------------------------------------------------------------------------- |
| |
| ISSUE 5: CONTEXT NOT PASSED TO AGENT |
| ------------------------------------------------------------------------- |
| Symptoms: |
| * Agent doesn't see VA conversation |
| * Screen pop missing VA data |
| |
| Troubleshooting Steps: |
| 1. Verify Set Variable nodes in flow |
| -> Must set CAD variables before queue |
| 2. Check variable mapping |
| -> VirtualAgent.* variables must be mapped |
| 3. Verify Agent Desktop layout |
| -> CAD variables must be configured in layout |
| 4. Check transcript URL generation |
| -> Logging must be enabled in Dialogflow |
| |
| ------------------------------------------------------------------------- |
| |
| SUPPORT ESCALATION: |
| ======================================================================= |
| |
| If issues persist after troubleshooting: |
| |
| Cisco TAC (for WxCC issues): |
| -> https://mycase.cloudapps.cisco.com/case |
| -> Collect: Flow export, VA config, error screenshots |
| |
| Google Cloud Support (for Dialogflow issues): |
| -> https://console.cloud.google.com/support |
| -> Collect: Agent export, conversation logs, request IDs |
| |
+-----------------------------------------------------------------------------+
Document Summary¶
This appendix provides comprehensive step-by-step procedures for configuring AI integration in Webex Contact Center with Google Dialogflow CX.
Sections Covered: 1. Pre-Implementation Checklist 2. Webex Control Hub - AI Agent Configuration 3. Google Cloud Platform - Project Setup 4. Dialogflow CX - Agent Creation 5. Webex Contact Center - CCAI Connector 6. Flow Designer - Virtual Agent Integration 7. Agent Assist Configuration 8. Validation & Testing
Estimated Total Implementation Time: 4-6 hours (excluding testing)
For complete technical specifications and design rationale, refer to Chapter 10 main document.
© 2026 Abhavtech.com - Appendix 10-E v1.0