INTRODUCTION
Fireworks AI powers enterprise-grade AI deployment with MongoDB
Industry
Computer Software & Technology
Product
MongoDB Atlas
MongoDB Atlas Vector Search
Use Case
Gen AI
Customer since
2024
THE CHALLENGE
Supercharging generative AI models
Founded in 2022 by industry veterans from Meta’s PyTorch team, Fireworks AI has built a lightning-fast inference platform that curates, optimizes, and deploys over 40 different AI models. Fireworks AI can process 66 billion tokens daily while achieving four times faster inference speed and eight times higher throughput and scale than competing platforms.
Although large language models (LLMs) are powerful, they have limitations. They are probabilistic, which can lead to hallucinations. LLMs are also constrained by their training data.
Fireworks AI’s customers need a robust vector database that can provide proper context to models. These requirements led to a natural partnership with MongoDB, with the company joining the MongoDB AI Applications Program (MAAP). As a MAAP partner, Fireworks AI is helping organizations rapidly build and put applications into production with its high-performance inference platform supported by MongoDB.
“MongoDB is one of the most sophisticated database providers, and it’s very easy to use. We want developers to be able to use these tools, and we want to work with providers who enable and empower developers.”
Benny Chen
Co-founder, Fireworks AI
Co-founder, Fireworks AI
THE SOLUTION
Deploying enterprise-ready RAG with MongoDB and Fireworks AI
By partnering with MongoDB Atlas, an operational database with native vector search capabilities, Fireworks AI addresses one of AI adoption’s toughest challenges. This collaboration enables the secure integration of enterprise data with AI models, facilitating retrieval-augmented generation (RAG) while maintaining data sovereignty.
“We see RAG combined with LLMs all the time,” said Lin Qiao, CEO of Fireworks AI. “Fireworks AI is the industry-leading generative AI inference platform for LLMs, and MongoDB is the industry-leading vector search engine for RAG.”
Using MongoDB Atlas as its vector database, Fireworks AI offers:
- Retrieval-augmented generation that ingests vast document collections for enhanced conversational AI.
- Classification through semantic search for analyzing concepts across content types.
- Image-to-structured data extraction that transforms visuals into searchable formats.
- Real-time alert intelligence that automatically detects fraud and security threats.
MongoDB Atlas supports context retrieval from enterprise data, reducing LLM hallucinations while maintaining performance. By incorporating vector search capabilities into an operational database, MongoDB enables organizations to adopt this functionality without moving existing data to third-party systems.
Fireworks AI enterprise customers can use hosted APIs or deploy them into their own environments. With this setup, no data leaves the customer’s premises. This is important for industries like healthcare and financial services, for example, that have strict data sovereignty requirements.
THE RESULTS
Transforming AI deployment with enterprise-grade solutions
Using MongoDB Atlas with Fireworks AI improves model output quality and decreases latency, all at a lower cost. As AI infrastructure continues to mature, this partnership positions both companies at the forefront of enterprise-ready AI deployment. “Generative AI is going to be much more accessible,” said Qiao. “In the future, it will become more and more economical to use, to the point where it is going to be a utility.”
“Fireworks AI is the industry-leading generative AI inference platform for LLMs, and MongoDB is the industry-leading vector search engine for RAG.”
Lin Qiao
CEO, Fireworks AI
CEO, Fireworks AI
To learn more about building enterprise-grade, AI-powered apps with MongoDB, visit the MongoDB AI Application Program page.
Take the next step
Get access to all the tools and resources you need to start building something great when you register today.
.svg)