Case Study: VISUA achieves 10x scalable quality control with Qdrant

Visua and Qdrant Vector Search in Computer Vision

VISUA, a leader in computer vision data analysis for over a decade, faced significant challenges in scaling the quality control for its high-volume platform. Their existing method for identifying unclear object detection outcomes and anomalies was manual and inefficient, relying on sampling based on meta-information. To enhance its anomaly detection and reinforcement learning processes, VISUA sought a scalable vector database solution and evaluated several options, including Qdrant.

By implementing the Qdrant vector database, VISUA gained powerful hybrid query capabilities to combine payload data with vector search. This allowed them to rapidly deduplicate and identify similar images at scale, automating their quality control. The solution delivered a 40x increase in query processing speed and enabled the company to scale its quality assurance processes tenfold, significantly improving the accuracy of its computer vision algorithms through better data review.

VISUA

Alessandro Prest

Co-Founder