Case Study: Cambium speeds up national AI exam searches with MongoDB Professional Services

Cambium cuts MongoDB search latency from 55 minutes to 3 minutes with MongoDB

Cambium, an Israel-based software development company, faced severe performance issues with its MongoDB Atlas implementation for a national AI exam platform designed to detect cheating. As the launch date neared, the system experienced unbearably slow query times of up to 55 minutes and a high number of false positives, making the platform unusable for processing tens of thousands of exam papers.

MongoDB Professional Services provided the solution, working with Cambium to optimize the database configuration. MongoDB experts recommended vertical scaling to larger search nodes with more RAM and optimized the vector search parameters. This intervention resolved the latency issues almost instantly, reducing search times from 55 minutes to just 3 minutes, and also fixed the problem with false positives, allowing the critical national platform to launch successfully.

Cambium