Hussnain K.

I build AI systems that work in production, not demos, not prototypes. My focus is on three areas: LLM Agents and RAG pipelines, Computer Vision systems, and Document Intelligence workflows. I've delivered across construction, healthcare, and enterprise clients, architecting systems end to end from ingestion to deployment. Recent example: built an AI Constructability Review Agent for a construction firm that automatically reviews multi-page drawing sets, runs an 8-stage image preprocessing pipeline, detects MEP symbol issues using GPT-4o Vision, and flags missing tags, duplicate tags, and cross-sheet conflicts replacing several hours of manual senior engineer review per document set. MSc in Data Science and AI, University of Liverpool. 5 years building AI systems, originally from an embedded systems and automation background. Stack: Python, FastAPI, Docker, LangChain, LlamaIndex, FAISS, Pinecone, pgvector, OpenAI, Claude, Gemini, Llama, YOLO, OpenCV, PyTorch, TensorFlow. If your project needs production-grade AI, not a proof of concept. Send me the details and I'll tell you directly whether I can help and how.