Sightline AI launches VisionOps and raises $60M to bring observability to computer vision models

AI · 5 min read

Sightline AI unveiled VisionOps, a platform that instrumentizes computer vision models across edge and cloud to provide real-time metrics on drift, performance, and fairness. The company says VisionOps helps teams debug visual model failures using frame-level tracebacks and counterfactual visualizations.

The $60 million Series B was led by Aurora Capital with follow-on from existing investors. Sightline will use the capital to expand edge deployment tooling, build certification workflows for compliance-heavy industries, and scale its global support for on-prem installations.

VisionOps integrates with common CV stacks and includes alerting, automated retraining triggers, and a human-in-the-loop workspace for labeling ambiguous frames. Early customers in retail loss prevention and autonomous logistics report faster root-cause analysis and more maintainable models.