Company Description The Vydar Group develops advanced electronic navigation systems for NATO-friendly nations, with a focus on strengthening European technological autonomy. The team specializes in AI-driven navigation, tracking, and detection solutions for defense applications, including missile systems, aircraft, drones, and distributed sensing networks. Vydar builds its technology entirely in-house, from proprietary hardware to software and control systems, ensuring full control over performance, security, and supply chain with no reliance on restricted or foreign components. Team members work in a flat, collaborative, innovation-driven environment with direct access to the founding team and the opportunity to contribute to high-impact technology with real-world relevance in European defense.
As a Machine Vision Engineer at The Vydar Group, you will design, implement, and optimize machine vision algorithms for embedded systems deployed on constrained hardware in missile platforms, UAVs, drones, and distributed sensing networks. You will work closely with the founding team and technical partners to bring solutions from lab to field deployment, collaborating with hardware engineers on co-design and integration. Daily tasks include optimizing inference pipelines for memory, compute, and power constraints, training, quantizing, and deploying models using TensorFlow Lite and Keras, and benchmarking and validating vision systems against performance targets. You will maintain documentation and reproducibility of model training pipelines and contribute to continuous improvements in robustness and accuracy. You will join a team of 15to 20 highly skilled engineers with backgrounds in aerospace, microelectronics, embedded hardware and software, and AI-driven navigation. This is a full-time, on-site role based in Delft.
Qualifications
- BSc or MSc in Machine Vision, Computer Vision, Electrical Engineering, or a related field. MSc preferred, BSc considered with strong practical experience.
- Proven experience deploying vision models on constrained embedded hardware.
- Hands-on experience with TensorFlow Lite and/or Keras, plus familiarity with model quantization and hardware-aware optimization.
- Strong programming skills in Python and/or C/C++, and familiarity with relevant libraries such as OpenCV.
- Ability to debug complex system-level issues independently.
- Experience with ONNX/TensorRT, FPGA/MCU platforms, or defense, robotics, and AI projects is a plus. Startup or R&D experience is also beneficial.
- Good communication in English (Dutch is a plus), and the ability to collaborate in multidisciplinary teams and work on-site in Delft.
- NATO-friendly citizenship
