I build AI systems that hold up in production.

I'm an AI Engineer based in Riyadh with five years of shipping machine learning to production, first as a freelancer across 50+ client projects and since then inside product and enterprise teams. I work across the full delivery stack: data pipelines, experimentation, fine-tuning, inference optimization, deployment, and monitoring.

My recent work covers enterprise RAG and LLM assistants, on-premise model deployments for organizations whose data can't leave their infrastructure, speech-to-text pipelines, semantic search, and video intelligence.

Most of my modeling is in PyTorch and Hugging Face: LoRA/QLoRA fine-tuning of open-weight models such as LLaMA, Qwen, and Mistral, embedding and vision models for retrieval, and quantization and batching to make inference fast and affordable.

On the infrastructure side I use Docker, Kubernetes, AWS, and GCP Vertex AI, with MLflow, Weights & Biases, and GitHub Actions for tracking and CI/CD. I treat system design, from data versioning to serving and monitoring, as part of the job, not an afterthought.