rahul prasad prajapati

I’m a third-year B.Tech student at NIT Jamshedpur, focused on practical machine learning that goes from experiments to deployed systems. I work across classification, regression, clustering, and computer vision, and I’m comfortable turning models into usable products.

My strongest strength is building production-ready ML pipelines. I built and deployed a production-ready ML API using FastAPI and Docker, and I’ve packaged end-to-end workflows with scikit-learn tooling like Pipeline and ColumnTransformer.

I apply rigorous model improvement and evaluation in every project. I use Optuna for systematic hyperparameter tuning (including multi-objective studies) and I measure performance with metrics like accuracy, F1-score, ROC-AUC, RMSE/MAE, and confusion matrices.

I also enjoy modern AI applications—especially Retrieval-Augmented Generation. In my DocChat project, I built a PDF summarisation and Q&A chatbot using Gemini API, LangChain, and FAISS, then deployed it for real use.