ML Engineer, Computer Vision for Earth Observation

LiveEO builds satellite analytics products, including Twinspector, a very high-resolution satellite constellation. Its Sektion 4 team delivers government solutions and funded R&D projects from research through production.

• Design, train, and iterate on computer vision models for detection, segmentation, classification, change analysis, and geometric vision on very high-resolution satellite imagery
• Identify and adapt state-of-the-art computer vision and remote sensing approaches, moving from research papers to prototypes and validated baselines
• Optimize models for constrained and embedded targets, including on-device and onboard inference
• Build scalable training and evaluation pipelines across cloud and secure on-prem environments, with experiment tracking, reproducibility, and systematic failure analysis
• Deliver production-ready inference interfaces, model packaging, deterministic evaluation, and monitoring
• Collaborate with the data annotation function on labeling guidelines and edge cases, partner teams on validated deliverables, and external researchers; present findings clearly and efficiently

• Strong computer vision and machine learning fundamentals, including representation learning, supervision strategies, and evaluation design
• Practical debugging and optimization skills
• Experience across multiple computer vision problem types, with real depth in at least one
• Strong Python engineering fundamentals and deep experience with PyTorch, implementing and training deep learning models at scale
• Strong understanding of ML experimentation, versioning, and tracking
• Background in remote sensing, computer science, physics, a related field, or equivalent practical experience
• Comfortable working with researchers and presenting findings clearly and efficiently
• Willingness to work on-site in Berlin and eligibility to obtain a German security clearance (Sicherheitsüberprüfung)
• Будет плюсом: Experience deploying models under constrained compute or on edge/embedded devices; satellite or remote-sensing imagery experience; a PhD in a related field; Ray; Prefect or similar; AWS, secure on-prem or HPC experience, SLURM, Docker, DVC; GDAL, Rasterio, GeoPandas, STAC; PostgreSQL or similar; SAR, geospatial foundation models or VLMs; post-training VLMs (SFT, RLVR); synthetic data generation, sim2real or domain adaptation

• Flexible working hours and a hybrid work model
• Overtime is limited to necessary cases and offset with time off and rest
• Office in central Berlin Kreuzberg, with free fruit, nuts, and drinks
• Potential participation in the employee stock option program
• Urban Sports membership, BVG subsidy, and a corporate pension program
• Collaborative learning environment with internal workshops, knowledge-sharing sessions, journal clubs, and hackathons
• International environment with 30+ nationalities