Collaboration for Breakfast: MongoDB and Partners Share AI Insights at AWS re:Invent

Gregory Maxson

I’m old enough to remember when every tech conversation didn’t include the term “AI.” Hardly a day goes by without some mention of AI or generative AI (gen AI). But don’t just take my word for it: Google News search results for the phrase, “generative AI” have grown more than 2000% since ChatGPT was launched in November 2022.

The AI excitement is more than just hype. For example, we’re seeing widespread adoption of AI across MongoDB’s tens of thousands of customers. Meanwhile, a recent GitHub survey showed 92% of developers have already incorporated gen AI into their work, and Gartner predicts that by 2027, 90% of new applications will incorporate machine learning models or services.

MongoDB and our partners tapped into this excitement during AWS re:Invent. On November 29 — the same morning the company announced the integration of MongoDB Atlas Vector Search and Amazon Bedrock (and less than a week before Atlas Vector Search was made generally available) — MongoDB held an AI-themed breakfast that reinforced the importance of partnerships during this transformative time. During the breakfast, MongoDB product leaders sat down with leaders from four of the company’s partners — Gradient, LangChain, Nomic, and Unstructured—to share insights about building the next generation of AI applications.

Despite its 7 a.m. start time, the breakfast was packed with attendees from a range of industries and geographies — no small feat given re:Invent’s busy schedule — and excitement for AI was palpable. Given the broad interest in all things generative AI, organizations of all sizes want to learn how they can build the applications of tomorrow.

This is where MongoDB and partners come in: MongoDB provides an integrated developer data platform that accelerates innovation by simplifying the application development process. To streamline AI innovation, MongoDB partners with organizations that offer complementary technology solutions, interoperability, flexibility, and reliability.

Partnering to deliver a complete AI toolkit

For example, Unstructured works with MongoDB to help organizations connect enterprise data stored in difficult formats like PDF and PNG to AI models. And the combination of MongoDB and LangChain's application framework makes it possible to build solutions that leverage proprietary company data.

Meanwhile, with MongoDB Atlas Vector Search and Gradient, organizations can build, customize, and run private AI applications that leverage industry expert large language models (LLMs) to enhance performance. And last but hardly least, Nomic's tools allow users to visualize the unstructured data they store in MongoDB, to make AI more explainable and accessible.

All told each partner’s offerings work with MongoDB products to create a comprehensive set of tools with which developers can build AI applications.

At the breakfast, company leaders shared their thoughts on the current AI landscape, how their organizations collaborate with MongoDB, and what they see as the future of AI tools.

“At AWS re:Invent, we showed how MongoDB is the best platform for building enterprise-ready generative AI apps,” said Andrew Davidson, senior vice president of product management at MongoDB. “Our powerful developer data platform — which works seamlessly with cutting-edge AI ecosystem partners to enable openly composable architecture and design — empowers developers to create compelling AI apps and experiences with greater interoperability, simplification, flexibility, and choice, pushing the boundaries of what's possible.”

For example, LangChain Founding Software Engineer Jacob Lee noted that “it’s so, so early for generative AI. Most attendees at re:Invent had only just begun to consider principles and use cases for the technology. There is so much opportunity and potential impact yet to emerge that it will truly take the entire ecosystem's talents and creativity to explore it all.”

“In short, the most important thing is to support each other and just keep building cool things,” said Lee. Brian Raymond, founder and CEO of Unstructured, agreed that it's very early for generative AI. "We should start seeing incremental, yet exciting, gains in the performance of multimodal foundation models as well as increased focus on smaller models that are cheaper to run at scale," Raymond added. "It's likely going to take more time to mature the emerging foundation model stack (marked by retrieval-augmented generation) into a performant and cost-effective option for most organizations."

Creating a seamless AI development experience

Overall, the re:Invent breakfast conversation conversation highlighted how MongoDB and its partners are working together to create a holistic, seamless AI development experience.

By working closely with partner organizations to augment its industry-leading solutions, MongoDB ensures enterprises have access to everything they need in one place to develop cutting-edge, modern AI applications that are scalable, secure, and enterprise-grade.

“Gradient's mission is to democratize AI by making it more accessible to enterprises and developers,” said Chris Chang, CEO and co-founder of Gradient. “However in AI, data itself can be challenging which is why our partnership with MongoDB will allow users to make the most out of their data and leverage a best-of-breed technology to help power new AI features.”

To learn more about MongoDB’s artificial intelligence solutions—including resources to build next-generation applications — visit MongoDB for Artificial Intelligence. If your organization wants to build the next big thing in AI with MongoDB, consider applying for the MongoDB AI Innovators Program.