Senior ML Engineer (Computer Vision)

Role overview Own computer vision models from dataset preparation and training through evaluation, optimization, and production rollout. The work spans real time camera feeds and GPU or edge inference, with responsibility for model quality and the systems that make models practical to run.

Responsibilities - Develop and improve models for detection, segmentation, tracking, text recognition, license plate recognition, and face recognition. - Export and optimize models for production inference, including GPU and edge deployments. - Build evaluation datasets and metrics that reflect real world performance and reveal regressions before release. - Coordinate with platform engineers on model packaging, deployment, versioning, and monitoring. - Assess vision language approaches and use them where they outperform task specific models. - Guide technical decisions, review code and designs, and mentor other engineers.

Requirements - At least three years building machine learning systems, including computer vision models in production. - Expert Python and PyTorch skills, with strong NumPy and OpenCV experience. - Practical experience with modern object detectors and trackers. - Production inference experience with ONNX, ONNX Runtime, TensorRT, or Triton on GPUs. - Understanding of real time video pipelines, including RTSP, frame decoding, batching, latency, and throughput. - Comfortable working with Docker, Linux, and GPU environments; able to take ownership of outcomes and lead technical work.

Nice to have - Kubernetes experience. - Experience with vision language or multimodal models.

Benefits and work setup The role is described as remote, with substantial ownership in a small team and equity offered.