SHREE User

I’m a final-year B.Tech (AI & ML) student and Machine Learning Engineer focused on building production-deployed ML systems. I enjoy turning complex research ideas into reliable pipelines through strong analytical, logical problem-solving.

I designed a multi-agent LangGraph research pipeline (Researcher → Analyst → Writer → Critic) with dual-source retrieval using Tavily and ArXiv AI APIs. By combining multi-agent orchestration with structured, citation-backed report generation, I reduced research report generation time and improved retrieval accuracy.

I also fine-tuned Mistral-7B for Indian finance tasks using QLoRA (with PEFT) and tracked experiments in MLflow. I deployed fine-tuned adapters to HuggingFace Hub with live Gradio demos, taking models from preprocessing to end-to-end production deployment.

In my document work, I improved retrieval precision with hybrid search (FAISS + BM25) and CrossEncoder re-ranking, reducing irrelevant results in an offline RAG pipeline. I’m passionate about learning emerging AI technologies and applying advanced methods like LLM reasoning agents to build innovative solutions.