Harness Inc., a software delivery startup that provides artificial intelligence tools for developers to update and monitor applications, said today it’s shipping new features that help software teams orchestrate at AI speed.

In the past, a team of developers, software engineers, cloud architects and operations specialists could coordinate multi-service releases. Today, that has changed with the rapid drumbeat set by AI release cycles – separate tools and separate approvals for each part of the pipeline no longer make sense and slow down teamwork.

Harness argues that a unified solution, with AI at the heart of the DevSecOps workflow, is required for continuous delivery.

“AI has made it much easier to generate change, but that doesn’t mean organizations are automatically better at releasing it,” Intelliflo Ltd. head of DevOps Marc Pearce said. “The challenge now is coordinating changes across services, teams and controls without slowing everything down.”

The new features now include everything a team needs to coordinate multi-service releases, alongside AI-powered verification and rollbacks. This allows teams to decide rapidly when to proceed, pause or pull back features handled by automated verification.

The system also now includes a DevOps database management schema for Snowflake, allowing the database to change with code modifications smoothly. As more teams build atop data warehouse architectures, it can slow down high-level decisions when low-level code and table decisions must update to fit changing patterns.

Teams can now also experiment with warehouse-native feature management and governance, allowing them to roll out changes gradually to see how they affect their system. With experiments affecting the data and code, teams can fully understand how changes will affect their data management before they scale them.

According to Harness, the rush to build and launch code with AI has made the coding side cheap and fast. However, the release and verification part of the lifecycle can still be fraught with safety considerations; code still needs to be vetted, deemed safe for production and finally monitored in the wild. 

No plan survives contact with the customer, say the saying goes. In the end, moving fast and breaking things can lead to team burnout if their tools are causing them frustration. Harness said it wants to help DevOps teams build and deploy great software, and that means making the lifecycle boring.

Image: Pixabay