Yajat Suri

I’m a B.Tech CSE (Big Data Analytics) student focused on turning data and AI into practical products. I build full-stack systems with clean APIs, reliable persistence, and fast iteration—always with the user experience in mind.

In my “YouTube Learning Assistant” project, I created a full-stack AI platform that transforms videos into structured learning material. I built transcript extraction with the Supadata API, then generated executive summaries, detailed notes, and key takeaways using Gemini 2.5 Flash through prompt-engineered API calls. I designed REST APIs with FastAPI, persisted data with PostgreSQL (via SQLAlchemy ORM), and containerized backend services with Docker for scalable future features like “Chat-with-Video” and quiz generation.

I also developed a “Student Score Predictor” end-to-end ML pipeline using Pandas and Scikit-learn, performing EDA on 1,000+ records and engineering features across 5 categorical variables. I trained Linear Regression and Random Forest models (R = 0.845) and deployed an interactive prediction dashboard on Streamlit Cloud with real-time inference. I enjoy taking models from raw data to usable tools people can actually run.

Beyond building, I lead and coordinate—serving as Content Co-Head at NSUT-IIF and Operations Team Head in AI & Algorithms Society. I’ve curated event content strategies to drive outreach and improved recruitment messaging, scaling registrations from ~400 to 460+ (15% YoY). My technical direction is validated through achievements like the Goldman Sachs India Hackathon ranking (3,457/15,963 nationally) and my selection as a McKinsey Forward Fellow.