At Trust Innovation Limited, I contributed to TIL-EMS, which serves 10,000+ student records. I developed end-to-end features and worked on financial automation, transaction validation, and modularizing a monolithic module.
My AI projects include YouTube-Buddy, a Chrome extension that uses a RAG pipeline to answer questions about video transcripts. I built its retrieval layer with LangChain and deployed the Python backend on Azure.
I also built AI Interviewer, a voice interview platform with dynamic question generation and an evaluation pipeline, and Campus Haat, a student marketplace that uses image and voice AI features.
My accepted IEEE QPAIN 2026 publication benchmarks CoAtNet for multiclass lung cancer classification on CT images. I hold a B.Sc. in Software Engineering from Islamic University of Technology.
