Senior Computer Vision Engineer

Beschreibung

Job Zusammenfassung In dieser Position entwickelst du innovative Computer Vision-Lösungen, einschließlich Echtzeit-Objektverfolgung und 3D-Rekonstruktion, und bringst diese erfolgreich in Produktionsumgebungen. Dabei übernimmst du den gesamten Trainingsprozess sowie die technische Führung des Teams. Job Zusammenfassung In dieser Position entwickelst du innovative Computer Vision-Lösungen, einschließlich Echtzeit-Objektverfolgung und 3D-Rekonstruktion, und bringst diese erfolgreich in Produktionsumgebungen. Dabei übernimmst du den gesamten Trainingsprozess sowie die technische Führung des Teams. Deine Rolle im Team As Senior Computer Vision Engineer, you'll be instrumental in taking our automation platform from early prototypes into production systems deployed across logistics networks. You'll report directly to our CTO/Co-founder and work alongside a growing engineering team. This is a hands-on, high-ownership role for someone who wants to solve real-world problems at scale. You'll work on cutting-edge computer vision challenges real-time object tracking, multi-view 3D reconstruction, ID association, and behavioral anomaly detection in messy, industrial environments. Multi-camera fusion - Calibrate and fuse overlapping feeds into a single spatial model of the yard in real time. Tag-free re-identification - Associate the same asset across cameras using appearance and motion alone. 3D scene understanding - Map 2D detections to real-world coordinates and maintain queryable spatial state over time. Behavioral anomaly detection - Learned models that generalize across new sites, not brittle rule-based heuristics. Training stack - Own the pipeline from raw site footage to deployed model: labeling, experiment tracking, model versioning, and automated validation before anything ships to a customer site. Data flywheel - Design pipelines that turn live customer footage into training data at scale so model quality compounds with each new deployment. Edge deployment - Optimize for on-premise inference on commodity hardware. Production quality bar - Define, monitor, and own what 'good enough' means for mission-critical CV. Pilot site presence - Diagnose failures in the field. Technical leadership - Set standards for model development, code quality, and CV best practices as the team grows. Unsere Erwartungen an dich Ausbildung Advanced degree: MSc or PhD in Computer Science, Machine Learning, Robotics or adjacent. Qualifikationen You're excited about solving hard technical problems in real-world industrial settings. You communicate clearly, work autonomously, and thrive in a fast-moving startup environment. Strong expertise in Python and/or C++ and modern ML frameworks (PyTorch, TensorFlow). Deep understanding of computer vision algorithms, e.g. ByteTrack, DeepSORT etc. In-depth knowledge of CV tools and libraries: OpenCV, torchvision, YOLO, object detection/tracking frameworks. Fluent English (German not required). Erfahrung You are a highly skilled computer vision engineer with proven experience taking CV projects from research/prototyping into production. 5+ years of hands-on experience in computer vision, AI/ML or robotics projects. Production experience: you've taken at least one CV project from concept through deployment to end users. Cloud experience: hands-on work with at least one major provider (AWS, GCP, Azure). Code quality: experience with code reviews, testing, and maintaining large codebases. Unser Angebot Competitive salary + stock options (meaningful equity as an early engineer). Brand new office on Berlin's AI Campus (Merantix HQ) where you can collaborate with hundreds of AI engineers across the ecosystem. Hardware & equipment allowance to ensure you have everything needed to thrive. Build the core team with us. Benefits Gesundheit, Fitness & Fun 🎳 Team Events Themen mit denen du dich im Job beschäftigst Machine Learning Automation AI Computer Vision Metadaten Level: Senior Job Feld: Software, Data Anstellung: Vollzeit Vertragsart: Unbefristetes Dienstverhältnis Arbeitsmodell: Onsite Unternehmenstyp: Startup Branche: Internet, IT, Telekom Ort: Berlin | Deutschland

Tech Stack