Taking away the AI application heavy lifting
"We know that in order for a customer to take something to production, it has to be customised and fit their development environment"
Sponsored
It’s now the norm for people to use AI in their daily lives. But as people engage with AI, they rarely encounter externally-facing AI applications that are meaningfully company-specific or are built on company data.
“That's what companies are working to figure out,” explains Gregory Maxson, senior director of AI GTM (go-to-market) at MongoDB. “They all want to start using this technology in a way that's enriching their own products.
“It's something we talk to customers about every day.”
There’s no question that firms face data hurdles when working to build AI applications that work at scale. In such a fast-moving market, it’s a real challenge for businesses without robust data science experience to not only implement and integrate a full AI stack or run such applications confidently.
MongoDB takes out the friction
Through countless conversations with customers, MongoDB has worked to identify the friction points for customers and address them directly.
Today, MongoDB’s in-house expertise goes beyond just offering recommendations on the best models and tools for various use cases – optimised for performance, accuracy, latency and beyond.
MongoDB also has a wealth of internal experts to perform full stack implementations,integrations, and evaluations on behalf of customers.
“They don't have to do the code from scratch. We can give them something to get started,” Maxson tells The Stack. “Then we provide the guidance and expertise to help educate them on how to do this work themselves, so they can up-level their teams and be more confident running them.”
Rapid deployment
A central feature of MongoDB’s solutions suite is the MongoDB AI Applications Program (MAAP). Intended to help customers build enterprise grade applications, enhanced with AI technology, MAAP includes a series of accelerators that support customers throughout their AI journey.
Through quick start reference architectures that customers can rapidly deploy, MongoDB is equipping customers with the right tools to deliver value through AI applications. Built on top of MongoDB Atlas and a vetted, industry-leading ecosystem, customers can have a full stack application up and running relatively quickly.
“But we know that in order for a customer to take something to production, it has to be customised and fit their development environment,” explains Maxson. “We've assembled educational content as well as professional services to guide customers from these quick start guides to real production applications.”
From his experience, Maxson has seen how this full set of offerings is pushing customers over the finish line and helping them get from zero to one and then from one to 100 in their development.
“What's really cool is we're seeing customers transition from testing to launching these products out in the market, and seeing their customers get their hands on them and finding real proof of value. It’s all built on the data that they have stored with us and MongoDB.”
Rewriting the playbook
What Maxson enjoys most about his work at MongoDB is the fact that for many customers there’s not yet an obvious playbook, so he gets the opportunity to really think creatively about how to help customers.
In practice, this means bringing together an ecosystem of partners representing all the technology needs for building an AI app, ensuring customers have access to a variety of tools as well as professional services and expertise to integrate and implement these technologies. Some companies want to try new open source offerings, while others want to deploy a full stack of solutions that runs on, for example, Azure, GCP, or AWS.
“That's the great thing about MongoDB, we can run anywhere,” says Maxson. “We work with providers across lots of product categories, including AI frameworks, LLMs, embedding models, infrastructure and model hosting, and enterprise data connectors so customers can have all the components they need for a full stack AI application. Our customers have diverse needs and we want to meet our customers where they are and help them determine the right set of offerings that fit their business requirements.”
While all customers have unique needs, the efficacy of their applications, whether AI related or not, is paramount. Releasing an AI solution that could be wrong one in ten times is just not acceptable. With over 50,000 enterprise customers, MongoDB is a battle-tested, secure and scalable solution for customers who are looking to deploy production-grade AI applications.
“They need their data to be in a secure, scalable environment that runs anywhere that they want to run, on any cloud with any model,” concludes Maxson. “MongoDB offers both flexibility, as well as the security, which is proving to be incredibly valuable for customers as they move from this prototype phase of last year to the production phase of today.”