HCL Foundation · Public health learning
Public Health Graph RAG WhatsApp Assistant
Course knowledge becomes a conversational Graph RAG assistant on WhatsApp.
OutcomeGraph-backed public-health learning answers over WhatsApp
15 days · 2-person team
The Challenge
Public-health course content needed to be queryable through a conversational channel. Simple search was not enough because topics, concepts, and dependencies required relationship-aware retrieval.
The Results
- Graph RAGRelationship-aware knowledge retrieval
- WhatsAppAnswers delivered where users already work
- Graph-backed public-health learning answers over WhatsApp
- Graph RAG, Relationship-aware knowledge retrieval
- WhatsApp, Answers delivered where users already work
- 15 days, POC delivered by a 2-person team
What We Built
We partnered with HCL Foundation · Public health learning to ship a production-ready AI platform for Public Health · E-Learning. We built a content ingestion pipeline that structures learning content into a Neo4j knowledge graph. We combined graph retrieval with Pinecone vector search and OpenAI RAG for richer answering. We kept the engagement as a focused POC/build that proves Graph RAG delivery for public-health learning content.
How We Delivered
Built a content ingestion pipeline that structures learning content into a Neo4j knowledge graph.
Combined graph retrieval with Pinecone vector search and OpenAI RAG for richer answering.
Exposed the assistant through WhatsApp using Twilio for a two-way conversational flow.
Kept the engagement as a focused POC/build that proves Graph RAG delivery for public-health learning content.
Scope of Work
- Graph-RAG knowledge layer for public-health content
- WhatsApp delivery channel for field queries
- Grounded answers from curated health knowledge
- Operator-safe retrieval and response design
Tech Stack
Tools we deployed in production, not just POC'd.
Ready to build?
Tell us what you're trying to automate. First call is free and we'll tell you honestly what it takes.