Harness today added a vibecoding tool that extends the artificial intelligence (AI) capabilities of the Harness platform to enable DevOps teams via a natural language interface to automatically migrate databases
Stephen Atwell, a principal product manager for Harness, said this AI-Powered Database Migration Authoring addition to the Harness Database DevOps platform adds a vibecoding tool that, via a natural language interface,e enables application development teams to migrate database schemas and SQL scripts.
The AI-Powered Database Migration Authoring tool analyzes the current schema and policies, generates a backward-compatible migration, validates the changes made in a way that can be audited, and then commits those changes to a Git repository through which they can be managed via the Harness continuous integration/continuous deployment (CI/CD) platform.
The Harness platform then manages those changes much like any other code, where policies can be enforced and managed as code via a dashboard provided by Harness, said Atwell.
Trained using a set of database management best practices vetted by Harness, the AI-Powered Database Migration Authoring tool understands keys, constraints, triggers, backward compatibility, and compliance standards.
At the core of that capability is the Software Delivery Knowledge Graph and Model Context Protocol Server that is natively embedded within the Harness AI platform, rather than being a plug-in that has been bolted on, said Atwell.
It’s not clear how broadly vibecoding might be applied to DevOps workflows, but it’s apparent that vibecoding tools that understand intent described in natural language will soon be more widely embedded within DevOps workflows. As the transition occurs, it becomes simpler to extend DevOps workflows to databases, which have long been separately managed using manual processes that slow the pace at which applications can be built and deployed, noted Atwell.
Databases, while playing a critical role in the development of software, have historically been managed more by database administrators (DBAs) than DevOps engineers. However, since the launch of the Harness Database DevOps platform, it’s become more feasible to automate the management of databases alongside application code, said Atwell.
That capability also makes it more feasible for IT teams to incorporate DBAs into a larger platform engineering team, he added.
That’s also a crucial capability for IT organizations that are migrating databases from, for example, a proprietary platform such as Oracle to an open source alternative such as Postgres or a document database such as MongoDB, or conversely, converting a database used to develop an application into a set of Oracle databases that might be the internal standard for the production environment, noted Atwell.
Regardless of the direction of the migration, the overall goal is to provide IT organizations with more control and flexibility as changes are needed to be made to the underlying database platform, said Atwell.
The only issue left to resolve now is determining just how many different database platforms an organization may need to support as it becomes easier to migrate workloads from one to another.