Muhammad Fezan — Software Engineer

I'm a 6th-semester Computer Science student at Lahore Garrison University with a strong foundation in full-stack development, AI/ML systems, and agentic workflows — and a habit of building things that go well beyond coursework. My core stack is MERN (MongoDB, Express, React, Node.js) paired with Python/FastAPI for backend-heavy and AI-driven systems. I work extensively with RAG pipelines, multi-agent LLM architectures, and fine-tuning workflows (LoRA/QLoRA), with hands-on experience integrating models from OpenAI, Anthropic, and Mistral AI. Some things I've shipped: DataForge — AI-powered fine-tuning dataset generator with FAISS deduplication, SSE streaming, and multi-provider model routing (FastAPI + React/TS) LettaXRAG — Conversational AI combining RAG with Letta long-term memory for persistent, context-aware interactions TumorClassifier-RAW-vs-DIP — Brain tumor MRI classification pipeline comparing raw vs. DIP preprocessing with Linear SVM, OpenCV, React + FastAPI BTSC-UNet-ViT — Hybrid ViT + U-Net architecture for brain tumor segmentation, trained on ~90k images (submitted for publication review) Dr.Document — Multi-agent GitHub documentation generator using a CodeReader → RequirementsExtractor → Manager → FinalReviewer pipeline I'm particularly interested in the intersection of applied ML research and production engineering — building systems that are not just theoretically sound but actually work at scale. Always curious. Always building. 🚀

Stackforce AI infers this person is a Full-Stack Developer with a focus on AI and ML applications.

Location: Lahore, Punjab, Pakistan

Experience: 0 mo