I didn’t just build an AI chatbot. I built an AI system that knows where its answers come from. I...I didn’t just build an AI chatbot. I built an AI system that knows where its answers come from. I...
The network for creativity
Join 1.25M professional creatives like you
Connect with clients, get discovered, and run your business 100% commission-free
Creatives on Contra have earned over $150M and we are just getting started
Back to feedPost
Youssef's avatar
Youssef Talaat
• Sep 3
I didn’t just build an AI chatbot.
I built an AI system that knows where its answers come from.
I recently built a RAG-powered Clinical Evidence Copilot that combines:
→ Hybrid Retrieval → Vector Search → BM25 → ChromaDB → LLM Generation → Source-based evidence retrieval
Instead of asking an LLM to answer from memory, the system retrieves relevant evidence first and uses that context to generate a grounded response.
This approach can be used to build AI Knowledge Assistants, Document Q&A Systems, Internal Company Chatbots, Research Assistants, and Customer Support AI.
The retrieval and quality of context matter just as much as the LLM itself.
machinelearningAI Agent Engineerrag
Post image
Muhammad's avatar
Muhammad Sohaib
• 17h
Super impressive work
Back to feed
The network for creativity
Join 1.25M professional creatives like you
Connect with clients, get discovered, and run your business 100% commission-free
Creatives on Contra have earned over $150M and we are just getting started