Vision Systems
Computer Vision Engineer
You will help egly understand the world around the wearer without making the product feel invasive. The role is hands-on across model evaluation, on-device performance, datasets, privacy constraints, and production-quality vision features.
What you will build
- Develop hand, object, plane, scene, and motion understanding features for eyewear use cases.
- Optimize models and pipelines for latency, battery, memory, and thermal constraints.
- Build dataset, evaluation, and regression tooling for real-world indoor and outdoor conditions.
- Partner with design and privacy engineering so vision features are explainable and controllable.
What you bring
- 4+ years building computer vision, machine learning, robotics, AR, mobile camera, or embedded AI systems.
- Strong Python plus production experience in C++, Rust, Swift, Kotlin, or similar.
- Experience with PyTorch, Core ML, ONNX, TensorRT, MediaPipe, OpenCV, or equivalent tooling.
- Practical instincts around datasets, evaluation, failure modes, latency, and privacy-sensitive features.
Nice to have
- Experience shipping on-device ML for mobile, wearable, robotics, or camera products.
- Experience with SLAM, VIO, depth, segmentation, hand tracking, or object tracking.
- Experience with synthetic data, active learning, or model compression.
Remote setup
This role is fully remote. We work async-first with written specs, weekly product reviews, focused lab blocks, and planned travel for hardware bring-up, field testing, and team sessions when needed.
Process
Intro call, portfolio or technical review, practical work session, founder conversation with Marvin Huelsmann, and offer. We keep the process direct and aim to finish within two weeks.