A Node.js sample that wires Google Dialogflow into Webex Contact Center for AI-powered voice self-service: caller context flows from WxCC into a Dialogflow virtual agent, a Cloud Functions webhook fetches account or CRM data, and the result is spoken back to the caller — all without a separate middleware server. Covers both Dialogflow CX (recommended for new projects) and Dialogflow ES (for teams extending existing agents).

Why use this playbook

• Pick the right generation fast: A side-by-side comparison table in the README maps Dialogflow CX vs ES across runtime, dependency footprint, WxCC integration path, and "pick this if…" guidance so you decide before writing a single line.
• Start from working code: Both samples are adapted from official WebexSamples and show the exact request/response shapes WxCC and Dialogflow expect.
• Lighter where it counts: The CX sample uses only the Node.js built-in https module — no framework overhead — while the ES sample shows the full Firebase + dialogflow-fulfillment pattern for teams already on that stack.
• Secrets handled safely: All backend API URLs load from environment variables; no credentials are hardcoded.
• Easy to verify locally: Import the bundled WxCC flow JSON files and the mock API URLs let you confirm end-to-end behavior before connecting a real backend.

What it does

• Receives a Dialogflow CX webhook request and returns a fulfillment_response with an account balance fetched from a configurable backend API.
• Receives a Dialogflow ES fulfillment request, reads caller context (name, email, reason, pin) from originalDetectIntentRequest.payload, performs a CRM lookup, sets Dialogflow output context, and returns a personalized spoken greeting via the dialogflow-fulfillment SDK.
• Includes importable WxCC flow JSON for both variants so you can wire the demo end-to-end in Flow Designer without building flows from scratch.

Preview listing

This App Hub entry was published from an open pull request. Review the change.