At MongoDB.local NYC, the company revealed that Search and Vector Search are now in preview for both Community Edition and Enterprise Server. These features were previously limited to Atlas, MongoDB’s cloud service.
Before this release, teams running self-managed MongoDB often relied on outside vector databases. That added extra tools, more cost, and higher risk. Now, MongoDB is building the same tools directly into its core products.
Ben Flast, Director of Product Management at MongoDB, said the company started with Atlas to grow the service quickly. Now, after years of development, it is ready for broader use. The new design also lets the search engine run as a separate binary. That makes it possible to scale on its own, run locally, or place on different hardware.
With vector search, teams can add:
- Autocomplete and fuzzy search
- Faceted search
- Internal search systems
- Semantic search powered by AI
- RAG, agents, and hybrid search
- Text analysis
Early testing came from partners such as Volcano Engine Cloud, LangChain, and LlamaIndex.
MongoDB 8.2: Faster Queries and Stronger Encryption
MongoDB also launched version 8.2. It delivers big gains compared to 8.0:
- 49% faster unindexed queries
- 10% faster in-memory reads
- 20% faster array traversal
- Almost 3x faster bulk inserts for time-series
The update also improves Queryable Encryption. It now supports partial matches, which means text searches can run on encrypted data without exposing the raw information.
This makes it possible, for example, for a healthcare provider to find all patients with a diagnosis that contains “diabetes” while keeping medical records fully encrypted. In the past, teams often left sensitive fields unencrypted or built separate indexes just to enable queries.
MCP Server Reaches General Availability
MongoDB’s MCP Server has moved out of preview and is now generally available. It adds enterprise-grade authentication (OIDC, LDAP, Kerberos), proxy connectivity, and self-hosted remote deployment support. Teams can now share deployments and manage them with centralized settings.
MongoDB AMP: AI-Driven Modernization
MongoDB also introduced AMP, a platform that uses AI to speed up application modernization. AMP includes automation tools, a delivery framework, and expert engineers.
The AI agents handle tasks like adding documentation, writing tests, or migrating Java code to Spring Boot. MongoDB’s engineers step in for more complex needs, such as compliance or security requirements.
Many companies start modernization thinking they have one database, then discover many. AMP helps them deal with this hidden complexity, unify systems, and move toward a more standard setup.
Shilpa Kolhar, SVP of Product and Engineering, said automation makes modernization faster and reduces the number of people needed. She also stressed that legacy systems can no longer keep up with the speed of AI innovation.
The Bottom Line
MongoDB is expanding beyond its role as a database vendor. With Search and Vector Search for self-managed databases, MongoDB 8.2 performance upgrades, MCP Server GA, and the new AMP platform, the company is building a full platform for modern application development.
Resources:
https://sdtimes.com/
https://knowledgenexuses.com/
