Sortrace AB

We are looking for a Computer Vision Engineer to develop and improve our image-analysis models on an existing platform that is already in production. You will train, evaluate, and optimize deep learning models for real-world waste imagery, not research in isolation. What you'll do Develop, train, and optimize deep learning models for image analysis Evaluate and extend existing model architectures Improve accuracy, performance, and robustness Work with image datasets: processing, augmentation, and evaluation Experiment with architectures and training approaches Solve hard technical problems and iterate on production CV systems What we're looking for At least 3 years of hands-on computer vision and model development, or equivalent documented experience (professional work, open source, or serious personal projects) Experience building, training, evaluating, and optimizing DL models for images Solid understanding of CV architectures (e.g. CNNs, Vision Transformers) Strong Python PyTorch or TensorFlow in practice Real-world image datasets: preprocessing, augmentation, evaluation Ability to adapt and extend existing models against clear technical requirements We care about what you've built, not only job titles. Show projects, repos, or results you can point to. Nice to have Edge AI and running models on edge or embedded devices Optimization for limited compute (quantization, inference performance) Real-time image or video analysis Who you are Curious, solution-oriented, and happy to go deep on technical problems. You keep up with CV and DL, turn ideas into improvements, work independently and in a structured way, and collaborate well on technical decisions. What we offer Production computer vision with real impact A platform that already runs. You improve what ships. Room to influence architecture and methods Hard problems where precision, robustness, and performance matter