MongoDB for Artificial Intelligence
_Spot.svg?auto=format%252Ccompress)
The landscape of AI workloads
Generative AI-enriched applications
Today's smart apps go beyond predictive analytics to create completely new experiences — from chatbots offering personalized support to AI-generated images, code, audio, and video — all from natural language inputs and in real time.
Traditional AI-enriched applications
Most modern applications leverage analytics to create more meaningful and reactive customer experiences. Through the use of machine learning models, AI automates complex decisioning for personalization, fraud prevention, predictive maintenance, and more.
Feature store and inference store
Feature stores are systems used to store and serve attributes used in machine learning models. Offline feature stores prioritize high throughput and are used for model training. Inference stores, or online feature stores, prioritize low latency for real-time inferences and serving live applications.
Resources for building AI-powered applications
Discover how to leverage MongoDB to streamline development for the next generation of AI-powered applications.
View ResourcesCore platform capabilities
Unify diverse data services
Simplify the AI lifecycle through operational, analytical, and AI data services that leverage a single data model and single query API on top of a highly scalable and secure multi-cloud platform.
Flexible document model
Innovate and experiment with new parameters and data of any type by landing, storing, and indexing data without lengthy schema design or ongoing modifications.
Optimized storage and tiering
Achieve high throughput and low latency for inference stores and by combining data tiering and federation with row and column indexing in a horizontally scalable, operational database.
Expressive, familiar API
Enhance productivity for developers and ML/AI teams with a single expressive query API that simplifies data preparation, model training, inferencing, and knowledge retrieval.
Native vector capabilities
Augment applications with generative AI through natively integrated vector and document datastores without the extra infrastructure to provision, secure or manage.
Extensive integrations
Build AI-enriched applications with the leading multi-cloud developer data platform and a robust AI partner ecosystem including MLOps platforms and open source LLMs.
“All of the business logic of our platform needs to make decisions very fast. It’s all running on Atlas, which allows us to move faster because we don’t have to worry about deploying or scaling. Whenever we need to add more fields, more queries, more steps in the business logic, Atlas just works. We don’t have to think about it.”
“With Atlas Vector Search now we have a comprehensive vector/metadata database – that’s been battle-tested over a decade – that solves our dense retrieval needs [to perform retrieval-augmented generation]. No need to deploy a new database we have to manage and learn.”
Build ML/AI applications with a suite of integrated data services
Database
A multi-cloud database service built for resilience, scalability, data privacy, and security.
Vector Search
Unified with Atlas Database and with support integrations into LLMs, Atlas Vector Search is a fast and easy way to build semantic search and AI-powered apps.
Full-Text Search
Combining keyword search by Atlas Search with semantic search powered by Atlas Vector Search to improve the relevance and accuracy of prompts for LLMs.
Triggers and Functions
Automatically run code in response to database changes, user events, or on preset intervals. Easily interact with ML models deployed as REST endpoints.
Time Series Collections
Build and run data-intensive analytical applications by combining the flexibility of the document model with time series collections.
Data Federation
Seamlessly query, transform, and aggregate data across Atlas databases and cloud object storage.
Atlas Stream Processing (Preview)
Transform building ML/AI apps requiring skew detection, feature stores, and enrichment pipelines. Unify working with data in motion and at rest.
Spark Connector
Build new classes of sophisticated AI apps combining MongoDB data and high volume, high-velocity data in Apache Spark and Databricks.
PyMongoArrow
Efficiently move data between MongoDB and leading ML libraries including Pandas and Scikit-learn.
Application-Driven Intelligence: A guide for technology leaders
To compete and win in the digital economy you need to make your applications smarter. Smarter apps use data, AI, and analytics to engage users with natural language, generate insights and autonomously take action.
To build this new generation of apps, we need to do things differently. We can no longer rely on just copying our data out of operational systems into centralized analytics systems. Instead, we have to bring a new class of AI and analytics processing directly to the source of the data – to the applications themselves. We call this application-driven intelligence.
Defining the next wave of modern apps
MongoDB Atlas puts powerful AI and analytics capabilities directly into the hands of developers in ways that fit their workflows, frameworks, and languages.
Learn more about the requirements to successfully deliver application-driven intelligence and how you can get started.
MongoDB AI Innovators Program
Go from 0 to 1 to enterprise-ready with MongoDB Atlas and LLMs
Get started with MongoDB Atlas, popular generative AI frameworks, and LLMs to build your next-generation application that can perform at scale
Turning MongoDB into a Predictive Database
Using MongoDB and MindsDB to enhance predictive capabilities for data science and data engineering teams.
Semantic Search Tutorial with Atlas Vector Search and OpenAI
Learn how to make a call to an OpenAI API and perform a vector search query in MongoDB Atlas.
Use MongoDB Atlas and Databricks Together
A deep dive on integrating MongoDB Atlas and Databricks.
Build an ML-Powered Underwriting Engine
Creating a usage-based insurance model using MongoDB and Databricks.
_Spot_BS_SpringGreen.svg?auto=format%252Ccompress)