2 min Devops

MongoDB makes services available to build enterprise-specific AI tools

MongoDB makes services available to build enterprise-specific AI tools

MongoDB has made its dedicated AI services Atlas Vector Search and Atlas Search Nodes generally available. These services should help developers more easily develop and deploy new AI applications.

With the general availability of MongoDB Atlas Vector Search, the cloud-based database provider aims to help developers easily, quickly and securely integrate generative AI and semantic search functionality into applications. This helps users leverage the company’s operational data.

Integrating Generative AI

This service supports generative AI applications by providing a vector database with high performance and scalability. This is integrated with a global distributed operational database that houses all enterprise data. Using a single API, developers can easily build their own generative AI applications for any type of workload that can reside in the infrastructure of multiple major cloud providers. All without the non-essential data duplication and synchronization that other vector databases require.

The service additionally integrates with the AI models of multiple developers. Among them are OpenAI, Hugging Face, LangChain. LlamaIndex, Cohere and Nomic are mentioned.

MongoDB Atlas Search Nodes

MongoDB’s second now generally available service, Atlas Search Nodes, provides a dedicated infrastructure for applications that use generative AI and relevance-based search functionality. This serves to scale workloads independently of the underlying databases and support high-throughput cases.

Een diagram van een verkeersmodel en een diagram van een verkeersmodel.

The infrastructure environment optimizes to deliver the required computing power for workloads while fully scaling search requirements. This can be useful for MongoDB Atlas Search and Vector Search, among others. Searches can be up to 60 percent faster, according to the cloud-based database specialist.

‘Memory optimized’ option

With the general availability of the Atlas Search Nodes infrastructure in public preview, a ‘memory optimized’ option for low processor utilization has now been included. This option is especially suitable for using MongoDB Vector Search in production environments.

MongoDB Atlas Vector Search is now generally available through AWS, Google Cloud and Microsoft Azure cloud environments. MongoDB Atlas Search Nodes is currently available only through AWS. Availability for Google Cloud and Microsoft Azure will follow soon.

Also read: VMware Tanzu offers enhanced support for building AI apps