All projects
AI & LLM Agents
Production-Grade RAG System
Retrieval-augmented generation with hybrid search, semantic caching and RAGAS-based evaluation.
- Role
- Developer
Overview
A RAG pipeline designed like a production service rather than a demo:
- Hybrid search (keyword + vector) for better recall
- Semantic caching to cut latency and LLM cost on repeated questions
- Systematic quality measurement with RAGAS (faithfulness, answer relevance, context precision)
Repository link will be added when the code is published.