AI NewsInfrastructureAnnouncement

MongoDB buys Firebolt to add real-time analytics to Atlas

MongoDB has acquired analytics engine Firebolt. It plans to run the engine next to Atlas so agents can query live data.

AI News

Editorial2 min read

LinkedInX
MongoDB brand card for the Firebolt acquisition announcement.

Image: MongoDB

Why it mattersTeams building AI agents that read and write the same database get one store instead of two. The warehouse copy goes away.

A data warehouse is a copy of yesterday. MongoDB has bought Firebolt, an analytics engine built for live operational data. MongoDB plans to run it next to its own database. The combined platform is for applications and AI agents that act on what is happening right now.

The announcement is on MongoDB's own newsroom. It does not give a price. MongoDB says the engine will sit beside Atlas, its managed database service. The two will sync on their own. Customers will be able to switch real-time analytics on "in the coming months".

What Firebolt brings

Firebolt is a cloud analytical database. It is built for short, high-concurrency queries over large tables. Its chief technology officer, Mosha Pasumansky, said on the MongoDB page: "Firebolt was designed for the cloud to do what many people believed was impossible: low-latency, high-QPS analytics with fast-changing data."

MongoDB cites one named customer that runs it this way. Similarweb uses Firebolt over "more than a trillion rows at petabyte scale". The company that is being acquired describes sub-second response times on fresh data. These are MongoDB's numbers, in a post that announces the deal.

Mosha Pasumansky's past work is on the page too. He led Microsoft Analysis Services. He later worked on Google BigQuery. CEO Benjamin Wagner was the researcher on the earlier project that led to Firebolt.

Why MongoDB is doing it

MongoDB says most useful analytics have moved out of the dashboard and into the application. The post gives one example. A payment platform has to spot an unusual pattern before the transaction clears. The morning after is too late. A separate warehouse is a lagging copy of the same data.

MongoDB's chief executive, Dev Ittycheria, said in the post that Firebolt removes the "friction and latency" of a separate pipeline. The company has made several moves this year to keep data inside Atlas. It bought embedding provider Voyage AI. It released an Atlas Agent Engine for building agents with memory. It is now adding analytics on top. All three sit in the same product.

For a team building an agent that reads and writes to one store, this is a different shape of database. One store does the write, the read, and the analysis. Is this cheaper than the two or three pipelines teams run today? Each buyer has to answer that on their own data. The terms of the deal are not public, so the price of that convenience is also not public.

Source

MongoDB Acquires Firebolt to Bring Lightning-Fast Real-Time Analytics Into Its Intelligent Data Platform, MongoDB.

SourceMongoDB

This item was written by an AI system from the linked source. Reveneau is responsible for what it publishes.

Share
LinkedInX
Start a project