Computer Vision & MLOps Engineer
Tata Consultancy Services (TCS) | Client: PostNord AB · Sep 2023 – Present
- 1.
- Designed, trained, and fine-tuned YOLOX object detection models, improving detection accuracy (mAP) by 15% across diverse logistics environments.
- 2.
- Optimized production inference pipelines using ONNX and TensorRT, reducing GPU utilization by 40% while sustaining 30+ FPS across 600+ camera streams.
- 3.
- Developed Flask/FastAPI microservices with WebSocket streaming capable of processing 500+ real-time detection events/sec integrated with Azure Event Grid and Azure SQL.
- 4.
- Deployed containerized AI applications on Docker, Kubernetes, and OpenShift with Horizontal Pod Autoscaling and CI/CD, reducing deployment cycles from days to hours.
- 5.
- Owned the complete ML lifecycle including data preparation, model training, optimization, deployment, monitoring, debugging, and production support.
- 6.
- Collaborated with operations, QA, and DevOps teams to deliver scalable production AI solutions.

