Kognic
7 Case Studies
A Kognic Case Study
Zenseact, Volvo Cars' autonomous driving software partner, faced significant scaling issues with their manual data annotation pipeline. The process was slow, expensive, and struggled with quality consistency across different teams. Their engineers were spending too much time on manual quality checks instead of product development. To address this, they partnered with Kognic to implement a pre-label workflow and Auto-QC automation.
Kognic's solution involved a pre-label workflow where model-generated annotations were verified by humans, and an Auto-QC system to catch errors early. This dramatically cut annotation time per sequence and boosted throughput for Zenseact. The partnership saved engineering review time and provided transparent cost control, leading to savings of 10.5 MSEK through pre-annotations. Kognic helped Zenseact scale high-quality annotations while maintaining the safety standards required for production vehicles.