Ruturaj Bhosale

I’m a B.Tech (Computer Science and Engineering) student focused on building practical GenAI solutions—especially RAG systems that are fast, safe, and measurable. I’ve worked on AI projects that combine retrieval quality, latency-aware design, and guardrails for real-world usage.

In my AI internship (Jan 2025–Feb 2025), I built a health assistant chatbot using Python, LangChain, and the Pinecone vector database. I added conversational memory and metadata-filtered search, implemented dynamic query routing between general and medical databases, and deployed it as an interactive Streamlit app with robust safety and input guardrails—reducing latency by 30%.

On my projects, I aim for production-grade reliability. I developed a role-based, compliance-aware RAG engine with document access-control policies and cache-aware prompt formatting to reduce token cost, plus safety gates to detect prompt injections and enforce output compliance.

I also built an automated RAG evaluation framework in Python using Precision, Recall, MRR, and NDCG, along with model-graded LLM-judge evaluation (Faithfulness and Relevance). Beyond engineering, I’ve led frontend development for a society website using React and mentored 10+ juniors, and I’m motivated by impact-focused achievements like WorldQuant (Alphathon Silver) and Science/technology/society (Gold).