About me
I am a machine learning engineer in Abu Dhabi building production machine learning, computer vision, NLP, LLM, and data systems. My work spans edge inference, privacy-aware vector search, document intelligence, high-scale video understanding, and cloud data platforms.
I like taking models from research notebooks into services people can actually use: deployed APIs, optimized inference with NVIDIA Triton, TensorRT, OpenVINO, Docker, Kubernetes, and CI/CD, and measurable outcomes across enterprise and research settings.
What I'm building
-
Production ML systems
Deploying and optimizing models with Triton, TensorRT, OpenVINO, Docker, Kubernetes, Azure ML, AWS, and CI/CD.
-
Computer vision research
Building robotics perception, remote sensing, segmentation, object detection, and 3D point cloud systems with first-authored research at IROS and ECCV workshops.
-
LLM and vector APIs
Designing FastAPI and Flask services with embeddings, FAISS, Milvus, Redis, MySQL, MongoDB, LLM chat flows, and RESTful CRUD patterns.
-
Data platforms
Engineering analytics and prediction systems with Spark, PySpark, SparkSQL, Azure Databricks, PostgreSQL, Jupyter, and big data pipelines.
Selected impact
-
High-throughput inference
Deployed and optimized production machine learning APIs with NVIDIA Triton, TensorRT, OpenVINO, Docker, Kubernetes, and CI/CD.
-
Document intelligence
Trained vision and language transformer systems for key information extraction from semi-structured documents such as invoices and receipts.
-
Large-scale media ML
Built computer vision classifiers and segmentation services for high-scale video and collaboration products.
-
Security for autonomy
Researching adversarial robustness and object detector behavior to improve perception systems in safety-critical environments.
