MongoDB Inc. (NASDAQ: MDB) announced new artificial intelligence capabilities at its London conference, introducing features designed to help enterprises deploy AI agents in production environments.

The company released MongoDB 8.3, which delivers up to 45% more reads, 35% more writes, 15% more ACID transactions, and 30% more complex operations compared to MongoDB 8.0. The update requires no changes to application code.

MongoDB introduced Automated Voyage AI Embeddings in MongoDB Vector Search, now in public preview, which automatically generates embeddings as data is written or updated. The company stated its Voyage AI embedding models rank first on the Retrieval Embedding Benchmark.

The LangGraph.js Long-Term Memory Store became generally available, providing JavaScript and TypeScript developers with persistent, cross-conversation agent memory powered by MongoDB Atlas.

"The hardest part of running agents in production isn't the model. It's the data layer underneath it," said CJ Desai, chief executive officer of MongoDB. "To trust an agent at scale, it has to retrieve the right context, hold memory across sessions, and operate at machine speed."

MongoDB also announced cross-region connectivity for AWS PrivateLink is generally available, ensuring database traffic between MongoDB Atlas clusters in different AWS regions remains on the AWS private network.

Additional releases include a Feast Feature Store integration with MongoDB, new query expressions for data transformation, and MongoDB AI skill badges for developers.

The announcements were made based on a company press release.