At Technogram, I built and tested RAG pipelines with vector databases and LLM APIs for semantic search and document question answering. I also integrated AI models into backend services through REST APIs and collected and prepared datasets for machine learning.
In my projects, I fine-tuned BERT on SQuAD for question answering and developed a RAG system that answers questions from PDF documents using LangChain, FAISS, and Google Gemini. I’m studying for a BSc at Jahangirnagar University.

