MongoDB, Inc. announced new capabilities at MongoDB .local London 2026, furthering its vision and strategy of delivering a unified AI data platform that gives enterprises everything they need to run agents in production—a real-time database, full text and vector search, memory, embeddings, and reranker models—all in one platform. Until now, enterprises have had to stitch together disparate systems and to hope they would work together at scale. MongoDB has solved that.
“The hardest part of running agents in production isn’t the model. It’s the data layer underneath it,” said Mr. CJ Desai, President and 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, wherever the enterprise needs it. That’s why AI-native companies like ElevenLabs build voice agents on MongoDB, and why institutions like Lloyds Banking Group trust it for mission-critical workloads.”
Retrieval accuracy
With Automated Voyage AI Embeddings in MongoDB Vector Search, now in public preview, embeddings are now generated automatically as data is written or updated to give agents accurate, real-time context.
Agents are only as good as what they remember and what they can retrieve. Embedding models convert information into vectors—an array of numbers that represent meaning mathematically—so an agent can find the right information. MongoDB’s Voyage AI embedding models rank #1 on the Retrieval Embedding Benchmark (RTEB). This means agents built on MongoDB can accurately find the right information.
Automated Voyage AI Embeddingsremoves the manual infrastructure work that has historically stood between enterprises and accurate AI search. Enterprises that previously spent weeks building search infrastructure can now ship semantic search in minutes.
High accuracy requires strong memory. Agents without memory can’t learn, improve, or be trusted. The LangGraph.js Long-Term Memory Store, now generally available, gives JavaScript and TypeScript developers persistent, cross-conversation agent memory that Python developers have had—powered by MongoDB Atlas, as a single backend, with no additional database required.
“When AI tools and agents produce a wrong answer, the instinct is to blame the model,” said Mr. Pablo Stern, Chief Product Officer, AI and Emerging Products at MongoDB. “But the data platform is what enables the agent with the right context and memory to act correctly. With MongoDB, we’ve made this easy. Developers no longer have to build and maintain data infrastructure, wire up embeddings, or manage syncing between systems. They can focus on business outcomes rather than the plumbing.”
Performance under pressure
MongoDB 8.3, available today, delivers up to 45% more reads, 35% more writes, 15% more ACID transactions, and 30% more complex operations over MongoDB 8.0—without changing a line of application code.
When enterprises like Adobe need to scale to serve Fortune 500 marketing teams on one of the world’s most widely used platforms, the requirements are clear: sub-100ms retrieval, sub-second context updates, and zero downtime. MongoDB Atlas is built for AI speed.
“The requirements of enterprises running AI at scale are what we build for. MongoDB 8.3 makes agent workloads faster and cheaper to run on infrastructure customers already have. We’ve also moved common data transformations into the database itself, so teams no longer have to maintain external pipelines just to feed their agents. Production AI doesn’t wait, and neither do we,” said Mr. Ben Cefalo, Chief Product Officer, Core Products at MongoDB.
Run anywhere
For banks, healthcare organizations, and government agencies, deployment choice isn’t optional. It’s often a data residency requirement set before architecture enters the conversation.
MongoDB runs across Amazon Web Services (AWS), Google Cloud, Microsoft Azure, on-premises, and in hybrid environments. Customers get one database, one API, and one set of skills that work consistently wherever they deploy.
Cross-region connectivity for AWS PrivateLink, now generally available, ensures that database traffic between MongoDB Atlas clusters in different AWS regionsstays on the AWS private network, with no exposure to the public internet. That helps security teams approve cross-region architectures faster, with fewer exceptions, and without forcing a tradeoff between compliance and global reach.
With these announcements, MongoDB continues to deliver what enterprises need to run AI agents in production—all in one platform.
If you have an interesting Article / Report/case study to share, please get in touch with us at editors@roymediative.com roy@roymediative.com, 9811346846/9625243429
