WhatsApp AI Support Integration

I need my WhatsApp Business account wired up to the OpenAI API so every new message receives an instant, AI-generated support reply. The assistant’s answers must be drawn from our existing knowledge-base articles, not from canned scripts, so customers always get up-to-date and brand-consistent information. Scope • Build or configure the WhatsApp API endpoint (Cloud API, Twilio, or another compliant gateway—whichever you recommend and can set up end-to-end). • Create the middleware that forwards each incoming chat to OpenAI, injects relevant knowledge-base content, and returns the model’s response back to WhatsApp automatically. • Map typical customer-support intents (order status, returns, troubleshooting, etc.) to the knowledge-base sections so the model has the right context before it generates an answer. • Implement a simple fallback: if confidence is low or the question falls outside the knowledge base, escalate to a live agent chat tag. • Provide a short admin guide (setup steps, environment variables, and how to add new articles). Acceptance Criteria 1. When I send a test question from WhatsApp, a relevant answer sourced from the knowledge base is returned within two seconds. 2. Messages outside the scope route to a live agent with a clear note. 3. Code and documentation are delivered in a Git repo and can be redeployed in my cloud environment with one command. Tech stack keywords you can freely use: Node.js or Python, WhatsApp Cloud API, Twilio, Flask / Express, OpenAI ChatCompletion, webhooks, Redis or PostgreSQL if you need caching, LangChain or similar retrieval helpers. Hand over the finished integration, documentation, and a quick demo video so I can see it working live in WhatsApp.

Project ID: 40414871