Amazon Web Services has extended the functionality of its database portfolio by releasing Amazon DocumentDB Serverless, a self-managed application that scales dynamically without requiring admin provisioning. The release will continue the original DocumentDB’s longstanding compatibility with MongoDB.
AWS’s DocumentDB was first released in 2019, and became a useful enterprise tool for enabling businesses to store, query, and index JSON data. In contrast with relational databases with their set schemas, the more flexible document databases allow varying document structures within a data collection. This makes them a good choice for use cases with rapidly changing data sets like gaming, social media, ecommerce, and web applications.
While the original DocumentDB is fully managed, with AWS handling routine tasks like patching, backups and monitoring, customers still needed to provision their own instances on the database. Now, DocumentDB Serverless advances the database’s self-managing features: the service automatically monitors activity and scales up—as high as millions of requests per second— and back down based on the ever-changing volume of requests. The database admin no longer has to plan resource management.
By AWS’s count, this serverless feature reduces costs by up to 90% when compared with manually provisioning a database for peak capacity. Certainly, serverless computing instances offer enormous benefit over manual provisioning. Deloitte research in 2024, based on Fortune 100 clients, found that serverless applications offer potential costs savings of up to 57%, though the saving rate can be highly situational.
Amazon touts the ability of DocumentDB Serverless to support deployments of AI agents. Agent workflows can be hard or in some cases impossible to predict, as an autonomous agent can select numerous paths to accomplish its assigned tasks. If, for instance, an agent is assisting with scheduling and setting up a corporate event for multiple executives, the bot may encounter a flurry of activity based on each staff member’s needs. A serverless document database can keep pace with this highly irregular pace of requests as they wax and wane over several weeks—all without requiring an admin to pre-plan compute resources.
G2 Krishnamoorthy, vice president of Databases at AWS, noted that agents have put unprecedented demands on databases. To support the AI agent market, “AWS has transformed how customers build and scale their applications by making database management truly effortless with serverless,” he said. A serverless database’s ability to automatically provision resources is “magnifying the importance of data as a differentiator for customers.”
AccelByte, a game development backend platform that has already deployed DocumentDB Serverless, used the AWS product to handle quick shifts in players. “Unpredictable player surges during a game launch or special event create massive scaling challenges,” said Tony Fu, vice president of Engineering at AccelByte. The serverless database helped the company deliver “a delightful experience to millions of players worldwide.” Also significant, “We can eliminate capacity planning for our database workloads and allow our engineers to focus on feature development.”
Amazon claims that its customers’ use of serverless databases has more than doubled over the last three years. That figure is in keeping with the now widespread adoption of serverless throughout large enterprise. Since AWS first launched serverless in 2014 with Lambda, the number of vendors offering serverless rapidly expanded to include all major hyperscalers. Precedence Research firm valued the global serverless market at $14.2 billion in 2024, forecasting that it will reach $54.3 billion by 2030.
