At Innomatics Research Labs, I built multi-turn GenAI applications using LangChain, Gemini API, and Groq, with session-based memory and context-aware response generation.
I developed RAG applications with ChromaDB and Hugging Face embeddings, testing prompts and retrieval behavior to ground LLM answers in retrieved context.
I deployed LLM applications through FastAPI, Streamlit, AWS EC2/S3, and Hugging Face Spaces. In my autonomous multi-agent RAG assistant project, I implemented PDF retrieval and a DuckDuckGo web-search fallback, then debugged agent state management and tool failures.
I also built a video-to-article pipeline that extracts YouTube transcripts and generates Medium-style articles, and an energy-consumption regression model with a test R² of about 0.95. I’m a recent B.Tech graduate in Computer Science Engineering, with a focus on AI and Data Science.
