Giradhar Gopal Kumar

I’m an Information Technology B.Tech student focused on building practical AI systems and data-driven products. I like taking complex problems—retrieval, ranking, and generation—and turning them into interactive experiences people can use.

In my projects, I’ve built recommendation and assistant-style applications end to end. Musify used a CountVectorizer + cosine similarity approach over 2,391 songs to generate top-5 suggestions, then integrated the Spotify API (via Spotipy) inside a Streamlit app and deployed it on Render with serialized model artifacts.

I also enjoy designing retrieval systems that feel conversational and reliable. In StackAssist AI, I implemented a RAG pipeline with LangChain, FAISS, and Google Gemini, using hybrid retrieval (FAISS semantic search + BM25 keyword search), multi-turn query rewriting, conversation memory, and a web-search fallback through DuckDuckGo—delivered through an interactive Streamlit UI with response modes and latency tracking.

I bring the same build-and-ship mindset to developer tooling and user-facing experiences. I’ve developed an AI-powered resume builder using the Groq API and Llama 3.3, with a React + Vite interface, template-driven outputs, instant PDF export, and live previews. Alongside this, I’ve been recognized in competitive and learning tracks—Top 1,500 in Google Big Code 2026, selected for Amazon ML Summer School 2025 from 60,000+ applicants, and a Flipkart GRiD 7.0 national semi-finalist—while solving 300+ DSA problems on LeetCode.