We are a team of experienced professionals with a long track record of building IT solutions, who look at complex systems in fresh ways using leading edge Artificial Intelligence. We discover hidden patterns and suggest the steps toward more efficient operations. We help our clients gain a competitive edge in...
18 листопада 2025
AI / ML — Embedded Vision Developer — C++ (вакансія неактивна)
віддалено
About us
We build real-time analytics that run directly on smart cameras and edge devices: object detection, tracking, classification, and distance/velocity measurement. We’re a small, hands-on team that ships reliable software where performance and clean engineering matter.
What you’ll do
• Develop and deploy on-device AI applications (native or containerized) on embedded Linux.
• Integrate object detection & tracking models.
• Build calibration tools (intrinsics/extrinsics, homography) for distance and speed measurement.
• Optimize models for embedded inference (TFLite/ONNX Runtime, quantization, profiling).
• Work with device SDKs/APIs to control video, optics/zoom, parameters, events, and overlays.
• Contribute to CI/CD, logging/metrics, and robust error handling.
Minimum qualifications
• Strong C++ and Linux development (memory, concurrency, performance).
• Experience with computer vision / AI inference (OpenCV + TFLite/ONNX Runtime or similar).
• Comfortable with Docker and cross-compiling for ARM.
• Ability to consume HTTP/REST APIs and learn new SDKs quickly.
• Clear communication, pragmatic debugging, and ownership mindset.
Nice to have
• On-device AI experience on smart cameras or other edge hardware (NPU/DSP/SoC).
• Working knowledge of camera geometry (intrinsics, extrinsics, homography) and basic tracking algorithms.
• GStreamer or low-latency video pipeline experience (RTSP, overlays, metadata).
• Model optimization: int8 quantization, performance tuning, (rotated) NMS.
• Implemented oriented bounding boxes (OBB), re-ID, or multi-object tracking.
• Experience designing an abstraction layer over multiple vendor SDKs.
How we work
• Lean, engineering-led team with quick feedback loops.
• Autonomy with responsibility for production quality.
What we offer
• Real product impact on edge AI—not just experiments.
• Opportunity to grow into technical leadership across platforms.
• Flexible schedule, remote-first culture, supportive team.
Apply
Send your CV/LinkedIn, a short note on a relevant project (edge AI, embedded vision, or SDK integration), and links to repos or demos if you have them.