“We’re using LLMs and AI to fully modernize old applications,” said Scott Sanchez, MongoDB‘s product marketing and strategy leader, in an interview at the Google Next conference last week in Las Vegas.
At the conference, the company introduced the integrated MongoDB cluster assessment as a menu item into Google’s Migration Center Use Case Navigator. This service provides a set of tools and best practices to move computer workloads onto Google Cloud.
Putting workloads on the cloud has long promised superior performance, reduced management and perhaps even costs savings. However, moving a decades-old application can be a daunting task, especially given the fact that probably the engineers who designed the system are no longer around (and probably left scant documentation).
Cloud migration has been a growing part of MongoDB’s business, Sanchez said, and generative AI has been helping with some of the more thorny cases. The fact that a schema can be generated on the fly as new data comes in, gives the NoSQL system the flexibility to easily adopt to other systems, and to easily ingest the findings of a generative AI data-producing tool.
The assessment offers a set of tools for the discovery of assets and planning of a migration from an existing on-premises MongoDB database — either the community or enterprise edition — onto the MongoDB Atlas hosted database service on Google Cloud.
With these tools, companies can save money, run things more efficiently and make better architectural decisions, the company claims.
For older deployments, the company also offers the Modernization Factory, a consulting service to root out technical debt when updating an application.
Bank App Modernization
Some MongoDB customers have already benefited from GenAI-boosted modernization efforts with MongoDB’s help. For instance, Swiss bank Lombard Odier was able to migrate code 50 to 60 times quicker than previous migrations and move applications from legacy relational databases to MongoDB 20 times as quickly.
The AI tooling automated repetitive tasks and Modernization Factory’s GenAI-based tools streamlined regression testing and automatically generated much of the code required.
Similarly, Bendigo and Adelaide Bank had reduced the time and cost of modernizing its core banking applications by 90% by using MongoDB’s modernization framework and AI-powered migration tools.
MongoDB’s Approach
Every modernization effort is different in its own way, Sanchez said. Most modernization consultation efforts begin by having a team come in with the AI tools to observe and document the system, including inspecting the code base, all the API calls and other interactions.
This data is used by the AI to build a test suite, with thousands or even tens of thousands of tests.
“Then we have the AI work backwards from these tests to build the most secure, fastest, most modern, MongoDB-powered application possible,” Sanchez said.
The customer can even elect to use multiple models for the design work, and can have the models compete against one another, and validate each other’s results.
“It’s a huge driver for us, because there’s so much data that’s locked up in many, many end-of-life platforms that people are afraid to touch for 100 reasons,” Sanchez said. “Or if they haven’t been afraid to touch them, they’ve been afraid of what the bill would be to call in a traditional professional services model, to have someone spend four years to modernize it.”
But as time goes on, the company will use AI to refine this process even further, he added.
“I think as the LLMs get better, as the code assistant tools get better, as we continue to invest more in the agentic framework, our agents will be learning, and our people will be learning with every customer we do this for.”
Google paid for the reporter’s trip and attendance of Google Next.
The post MongoDB Finds AI Can Help With Legacy System Migration appeared first on The New Stack.
MongoDB says that with these tools, companies can save money, run things more efficiently and make better architectural decisions.