The cloud database company MongoDB Inc. said today it’s partnering with Amazon Web Services Inc. to optimize Amazon CodeWhisperer, so it can provide superior suggestions that aid developers in application development and modernization.

CodeWhisperer is a generative artificial intelligence-powered coding assistant that was unveiled in June. It can generate code suggestions when prompted using natural language. Those suggestions are informed by the existing code within developer’s integrated development environments, but the collaboration with MongoDB will enable it to generate more superior code, the companies said.

MongoDB is the creator of the document-oriented MongoDB database, which is used for a wide range of data-intensive applications and prized for its ability to store information in multiple different formats. MongoDB Atlas is the cloud-hosted version of that database.

In September, the company announced a major update to Atlas, adding vector search capabilities that filter and aggregate unstructured data so it can be used to train generative AI applications more easily. With the update, the company positioned Atlas as the database of choice for AI training workloads.

Through its collaboration with AWS, MongoDB has provided massive volumes of curated training data for Amazon CodeWhisperer that’s aimed at MongoDB use cases. As a result, developers can now access enhanced suggestions through Amazon CodeWhisperer that reflect best practices, enabling them to ideate and create new features faster and accelerate the development of applications.

MongoDB cites a study by International Data Corp. that forecasts more than 750 million cloud-native applications will be built in the next two years. To meet this demand, developers are increasingly turning to generative AI coding assistants to boost their productivity, so they can focus on harder problems.

However, MongoDB says, many coding assistants are unreliable, because they lack access to high-quality code samples. The danger is that the code generated by such tools may not conform to best practices, increasing the risk of code being shipped with vulnerabilities.

This is what the companies say they’re trying to prevent, by training and evaluating Amazon CodeWhisperer on code and libraries specific to the MongoDB database. The result is “enhanced code suggestions” that developers can trust, allowing them to build much more quickly than before, they said.

MongoDB explained that it worked with AWS to train Amazon CodeWhisperer on highly curated content and code from MongoDB’s documentation, detailed use cases and common tasks, while applying the same best practices encountered by developers when working with MongoDB data. As a result, the companies said, CodeWhisperer can help developers more quickly write high-quality code when building data aggregations, performing database operations, and accelerating migration of applications to MongoDB for modernization.

Jared Reminer, chief technology officer of the managed services provider Cascadeo Corp., said his company uses CodeWhisperer along with Amazon SageMaker and Amazon Bedrock to deliver more consistent operations to customers and respond to their needs faster.

“Our developers choose CodeWhisperer for its optimized experience when working with AWS APIs, boosting their productivity and improving code quality when building on AWS,” he explained. “Through this collaboration to train and evaluate CodeWhisperer on curated MongoDB data, our developers can now get intelligent recommendations across both AWS and MongoDB Atlas, streamlining how applications get built and giving developers more time to focus on solving hard problems for clients.”

MongoDB said the enhanced CodeWhisperer suggestions are available for programming languages including C#, Go, Java, JavaScript and Python. They take advantage of features such as built-in security scanning and a reference tracker that provides information on the origin of each code suggestion.

Best of all, MongoDB said, Codewhisperer is free for individual developers with no qualifications or time limits, meaning its entire community of users can take advantage of the service today. All they have to do is install the CodeWhisperer extension to their integrated development environment of choice, provide an AWS Builder ID, and they’re good to go.

Andrew Davidson, MongoDB’s senior vice president of product, said generative AI has shown great promise in helping developers to build new applications. “With built-in security scanning and the ability to provide source and licensing information when suggestions resemble publicly available open-source training data, Amazon CodeWhisperer now provides developers building on MongoDB a unique experience that will get even better over time,” he said.

Image: AWS