Senior Machine Learning Engineer - Video Understanding

Scandit is hiring a Senior Computer Vision/Machine Learning Engineer in Italy to build video understanding technology for retail operations. The role focuses on action recognition in real-world store environments, where models must detect behavior reliably while meeting precision, recall, latency, and privacy needs. You would lead technical discovery and early development for a new product, turning uncertain requirements into experiments and production-oriented prototypes.

The work spans the full computer vision lifecycle: video data ingestion and preprocessing, annotation workflows, model training and evaluation, and deployment. You would select and adapt architectures, including vision transformers for temporal video understanding, and partner with product managers, engineering teams, and external stakeholders to set technical direction. Scandit seeks hands-on experience shipping commercial CV/ML systems, especially video action or behavior recognition; VLM fine-tuning and privacy-aware video pipelines are useful additions.

Highlights

• Lead technical discovery and early product development from 0 to 1.
• Build end-to-end CV/ML pipelines using real-world video data.
• Develop action recognition models for live retail environments.
• Shape technical roadmaps and annotation practices with cross-functional partners.