At VISWAM.AI, I built Python-based data pipelines for an AI corpus collection engine, adding data quality checks and structured validation. I also worked on preprocessing, normalization, tokenization, and debugging to improve data consistency for downstream workflows.
My research paper on hybrid deep learning for real-time IoT DDoS detection reported 99.41% accuracy on DARPA and 99.24% on KDDCup99. I’ve also built a RAG-based HR policy assistant with LangChain and a multi-agent wedding planner, and deployed a full-stack AI application using FastAPI, Docker, and GitHub Actions.
