I build production AI systems. Computer vision, MLOps, intelligent applications. Domain expertise that makes AI trustworthy in high-stakes environments.
Object detection, image classification, and visual inspection systems — from research and model training to GPU-accelerated production deployment with full observability.
End-to-end ML pipelines — model serving with TensorFlow Serving and gRPC, containerised deployment, Prometheus monitoring, Grafana dashboards, and production alert design.
Retrieval-augmented generation pipelines, LLM integration with streaming inference, vector databases, and intelligent applications built on top of foundation models.
Years inside safety-critical industrial environments means I understand what "reliable" and "auditable" actually require — and I build AI systems that meet those standards.
"Most AI engineers have never worked inside a safety-critical environment. Most industrial engineers have never trained a neural network. I have done both — at real scale, with real consequences. I build AI that is not just accurate but observable, auditable, and maintainable."