Embedded Computer Vision Engineer

At Ultralytics, we commit to relentless innovation in the AI space and seek team members who resonate with our ambition to produce the world's best YOLO AI models. If you're obsessed with AI, eager to make an impact on the world, and thrive in dynamic, high-intensity environments, we invite you to apply for a position on our team.

Hello, hola, 你好, こんにちは, नमस्ते, hallo, bonjour!

At Ultralytics, we’re on a mission to simplify AI for everyone
. As the creators of the world’s leading Ultralytics YOLO models, we empower millions of developers, researchers, and companies worldwide to build state-of-the-art computer vision applications with our open-source tools.

Following our $30M Series A round, we’re expanding rapidly across our global hubs in London, Madrid, and Shenzhen
. This is an opportunity to join a fast-scaling, high-performance team that’s redefining the future of vision AI — where ambition meets impact, and ideas become reality.

We move fast. We build boldly. We execute with purpose. And we do it together.

As Embedded Computer Vision Engineer, you’ll take YOLO models from PyTorch to production on edge hardware. You’ll build and maintain export and inference integrations in the Ultralytics package, helping models run accurately and efficiently across CPU, GPU, NPU, and SoC platforms.

You’ll own technical integrations with silicon and hardware partners, translating between model requirements and platform constraints. Working across export formats, inference runtimes, and real devices, you’ll ensure partners and their customers can deploy YOLO with confidence.

This senior role suits an engineer with strong ML fundamentals, hands‑on edge hardware experience, and the communication skills to work directly with internal teams and external partners.

• Build and maintain export and inference integrations, from model conversion to on-device support.

• Keep integrations working across Python, PyTorch, and vendor SDK releases.

• Review partner contributions and validate integrations on real hardware before release.

Deployment, benchmarking & accuracy

• Set up embedded boards and development kits for YOLO deployment and testing.

• Optimize models for target hardware, balancing latency, memory, and accuracy.

• Debug accuracy loss caused by conversion and INT8 quantization.

• Benchmark and profile YOLO tasks and model generations, publishing results in the documentation.

Partner integrations

• Own technical relationships with hardware and silicon partners, including active AMD work and integrations with Google and Nvidia platforms.

• Lead partner engineering syncs and define clear integration guidelines.

• Translate platform constraints into model, runtime, and deployment requirements.

• Work with partner compiler teams to support new YOLO architectures at launch.

Infrastructure, documentation & support

• Diagnose issues across models, runtimes, drivers, and devices with ML engineers and partner teams.

• Maintain hardware CI on self-hosted and partner-hosted devices, plus deployment Docker images.

• Write integration guides and device benchmark pages.

• Create developer content, live sessions, and demos for industry events.

• Support developers and customers through Git Hub, the community forum, and our sales team.

• Strong foundation in machine learning or computer vision, including how YOLO models work.

• Strong Python skills and working experience with PyTorch and ONNX.

• Hands‑on experience with TensorRT, OpenVINO, ONNX Runtime, or a vendor NPU SDK.

• Experience with INT8 quantization, calibration, and debugging accuracy loss after conversion.

• Hands‑on edge device experience with NVIDIA…