Mohammed Sufiyan

I’m a software developer building production AI systems across LLM chatbots, RAG pipelines, and real-time computer vision. At Alphabit Technologies, I designed and deployed a transformer-based LLM chatbot via FastAPI and structured prompt workflows, reducing average customer query resolution time by 45%. I also built a real-time CCTV inference pipeline using OpenCV, DeepFace, and YOLO, optimizing inference latency to under 120ms per frame and automating daily reporting for 500+ streams.

I’ve contributed end-to-end to both backend and user-facing experiences, including a FastAPI + React.js ISBN scanning tool integrated with the Google Sheets API, cutting inventory entry time by 80% and reaching 97% accuracy. From fine-tuning YOLOv8 models with PyTorch to exporting for edge inference on Raspberry Pi with sub-200ms response time, I focus on measurable performance, reliable deployments, and practical engineering.