Syed Muhammad Taqi

I’m a Software Engineer specializing in NLP and LLMs, and I enjoy turning language capabilities into real, usable products. I focus on fine-tuning, RAG (Retrieval-Augmented Generation), and the software architecture needed to deploy high-performance language models in real-world environments.

I’ve built multiple RAG-driven systems, including a RAG-based Quranic Q&A with secure authentication, semantic search, and an analytics dashboard. I also created a Medical RAG Assistant that delivers context-aware responses using specialized healthcare documentation.

I care about reliability and compliance in AI applications, so I built a Policy Compliance Checker using the Gemini API and FAISS vector search against CUAD datasets. I’ve also shipped end-to-end apps like a Streamlit-based food image classifier and developed NLP tools such as an article summarizer using T5 and Pegasus.

Alongside engineering, I bring teaching and communication strength from my role as a Teaching Assistant for Applied Human Computer Interaction. My approach blends clear modeling choices (transformers, sequence-to-sequence methods) with practical tooling (LangChain, Hugging Face, and PyTorch/TensorFlow) to deliver outcomes people can actually use.