End-to-end computer vision platform from annotation to model deployment
Supervisely operates a full-stack computer vision platform spanning annotation, training, and inference. The tech stack—Python, OpenCV, Docker, REST APIs with support for specialized formats like DICOM and LiDAR—reflects a mature ML infrastructure. Active projects center on synthetic data generation, multimodal pipelines, and domain-specific adoption (medicine), suggesting the company is moving beyond general-purpose annotation toward vertical specialization and automation tooling.
Supervisely provides an end-to-end platform for computer vision R&D covering image annotation, dataset management, neural network training, experiment tracking, and model deployment. The platform integrates custom and pretrained models, supports active learning and human-in-the-loop workflows, and includes automation tools like SmartTool to accelerate annotation cycles. Founded in 2017 and based in Tallinn, the company operates as a lean, Estonia-focused engineering organization with balanced hiring across data and engineering functions. Core customers span healthcare, autonomous systems, and enterprise AI teams.
Python, OpenCV, Docker, NumPy, Pandas, scikit-learn, REST APIs, Vue for frontend. Specialized support for DICOM, LiDAR, Waymo, and nuScenes datasets.
Yes. Active roles span data (4) and engineering (4) positions, predominantly mid-level and senior, all based in Estonia. Posting velocity is minimal.
Synthetic dataset generation tools, multimodal data pipelines, ML infrastructure, medical domain applications, and app-based platform extensibility for data operations and automation.
Other companies in the same industry, closest in size