Ali Hassan, an AI Engineer who ships past the notebook.

I'm an AI Engineer and Data Scientist working from Lahore, Punjab. Most of my week goes into agentic systems: agents that hold tools, hand work to each other and stay inside a latency budget. Right now that includes agentic voice chat agents, where the model has to listen, reason and answer inside a single turn.

The public record sits under alihassanml carries 150 repositories and 298 stars. The work spans multi-agent systems and voice agents built on LangGraph, Pydantic AI, MCP and local models through Ollama; retrieval-augmented generation, including federated and knowledge-graph RAG; computer vision, with YOLOv11 models for emotion, sign language, workplace safety and security; AI automation through n8n; and the MLOps that keeps any of it honest: Docker, AWS, DVC, CI/CD, experiment tracking.

The most-starred piece is a YOLOv11 face emotion detector at 31 stars, with smoking detection at 26, a self-driving steering model at 20 and 40-class sign language recognition at 12 behind it. I write mostly Python (72 of those repositories are Jupyter notebooks, 33 are Python), plus TypeScript when an AI product needs a front end people will actually use. Client work runs through Upwork, and what I hand over is a running system.