MongoDB Announces New Atlas Capabilities

Traditionally, MongoDB has held its signature event—MongoDB World—in June in New York City. MongoDB usually announces significant releases of their flagship database product at this event.  

However, this year, the company has transitioned to a global tour of “local” events that start in New York and then proceed to a few dozen regional events across five continents. 

And rather than announce a significant new monolithic release of the database product, MongoDB has revealed a grab bag of new capabilities instead.

Undoubtedly, MongoDB continues to show impressive growth that should please shareholders. MongoDB boasts revenue growth of 29% and a 40% growth in  Atlas cloud platform revenue. Atlas now accounts for 65% of total MongoDB revenue.   

Since the uptake of Atlas is so critical to MongoDB's overall revenue, we shouldn't be surprised that most of the new announcements are related to Atlas features. MongoDB announced five new capabilities in Atlas, and while none of these is revolutionary, each stand to contribute to further Atlas's growth.

The biggest news in our industry over the past year has been the sudden boost in the capabilities of generative AI using Large Language Models (LLMs). Atlas Vector search is a capability MongoDB claims will expedite those building applications using LLMs by allowing data to be encoded in a fashion that suits the machine learning algorithms central to machine learning routines. Vectors encode data to allow clustering and categorization algorithms to find semantically similar items.  

MongoDB also announced that Atlas Text Search would allow users to assign dedicated external nodes for better scalability and performance for vector and search workloads.

Atlas Stream Processing allows developers to create routines that work on high-velocity data streams. Streams may be consumed from Apache Kafka or Atlas Change Streams. The data consumed can be processed in real-time to raise alerts or events, which are then written to the Atlas database or back to Kafka.  

MongoDB also announced an enhancement to time series capabilities that allow data in time series to be updated and allowed the Atlas Data federation capability to support Azure, not just AWS.  

MongoDB used the NYC conference to announce several other developer-oriented capabilities, such as support for the AWS Cloud Development Kit—an Infrastructure as Code capability—better support for the Kotlin language, enhancements to the Atlas Kubernetes operator, and a Python library for analysing MongoDB data (PyMongoArrow). They also announced the general availability of MongoDB Relational Migrator, a tool that assists in migrating applications from relational databases to MongoDB.  

None of these new capabilities are particularly earth-shattering, but in composite, they represent a significant increase in the range of applications that MongoDB can address. 

Concentrating on these capabilities in the Atlas cloud platform may disappoint those using the community edition of MongoDB. Still, it makes sense commercially for MongoDB and represents increasing confidence within the company that they are primarily a vendor of cloud-based database capabilities. The cloud is the future for MongoDB, so that's where they are placing all their bets.