Enterprise Knowledge Assistant
Web + WhatsApp RAG Chatbot
AI · Enterprise Integration
Full-stack RAG development, indexing pipelines, messaging automation
Private production project
Overview
A full-stack Retrieval-Augmented Generation knowledge assistant that resolves queries in real time across web chat and the WhatsApp Cloud API, grounded in automatically indexed source cont[...]
The challenge
Enterprise knowledge is scattered across dynamic, JavaScript-rendered pages, and the people who need it are already on WhatsApp. The assistant has to index that content automatically and a[...]
Contribution
- Built a full-stack RAG knowledge assistant supporting real-time query resolution across web chat and the WhatsApp Cloud API.
- Developed automated Playwright-based indexing pipelines for extracting structured content from dynamic JavaScript-rendered sources.
- Engineered locally hosted LLaMA 3 inference workflows with webhook-driven messaging automation and admin dashboard analytics support.
Architecture & workflow
How the system moves from input to outcome.
Core features
Dual channel
One knowledge base answering on web and WhatsApp.
Real-time resolution
Queries answered as they arrive, not batched.
Dynamic indexing
Playwright handles content that only exists after JS runs.
Local LLaMA 3
Self-hosted inference workflows.
Webhook messaging
Cloud API webhooks drive inbound and outbound messages.
Admin analytics
Dashboard analytics support for usage visibility.
Technical highlights
- Playwright-based indexing of JavaScript-rendered content.
- Locally hosted LLaMA 3 inference workflows.
- Webhook-driven WhatsApp automation alongside web chat.