Case Study: Okta achieves faster, scalable identity search with MongoDB Vector Search on Atlas

Okta cuts operating costs by 30% with MongoDB

Okta, a leading identity security provider, sought a scalable way to power natural language queries for its Okta Inbox platform. Their initial approach required manually building and managing machine learning models and vector embeddings on dedicated infrastructure, which limited experimentation and scalability as the business grew. This created complexity and potential cost concerns for their development team.

By implementing MongoDB Vector Search on Atlas, MongoDB provided an integrated solution that automated embeddings storage and search within Okta's existing data platform. This allowed Okta's developers to focus on innovation and optimizing algorithms instead of infrastructure maintenance. The results included a smoother user experience, faster development iteration, and a projected 30% reduction in operating costs compared to their previous self-hosted solution.

Okta