B.E. CSE (Data Science) · Honors in AI & ML | RAG Pipelines · Multi-Agent AI · Generative AI
I build LLM-powered systems — RAG pipelines, multi-agent workflows, and document intelligence tools. Background in data analytics and business analysis; now focused on production-ready GenAI engineering.
| Project | Description | Tech | Project Live URL |
|---|---|---|---|
| Document Q&A RAG Pipeline | Production RAG system with Cohere reranking (top-20 → top-5), RAGAS evaluation, and Gemini 2.5 Flash streaming | Gemini, Pinecone, Cohere, RAGAS, Streamlit | https://ragdocumentassistance.streamlit.app/ |
| Onboard AI Agent | Production-grade multi-agent system with 3 chained Gemini 2.0 Flash agents automating the full employee onboarding journey — provisioning, 30-60-90 day learning path, and 24/7 HR Q&A | Gemini 2.0 Flash, Streamlit, Google AI Studio | https://onboard-aiagent.streamlit.app/ |
- Currently improve Multi Agent System with all the 4 agents interacting with each other
- Production RAG pipelines with embedding optimization, reranking, and automated evaluation (RAGAS)
- Improving LLM output quality through structured prompting, tool use, and agent memory