Dive Brief:
- AWS launched a group of frontier agents on Tuesday, which the company described as a “new class of AI agents that are autonomous, scalable, and work for hours or days without constant intervention.”
- The three frontier agents – Kiro autonomous agent, AWS Security Agent and AWS DevOps Agent – function as a virtual developer, virtual security engineer and virtual operations team member that can operate proactively and independently, according to the AWS announcement.
- “AWS DevOps Agent thinks and acts like a seasoned DevOps engineer, helping our engineers build a banking infrastructure that’s faster, more resilient, and designed to deliver better experiences for our customers,” Jason Sandery, head of cloud services at Commonwealth Bank of Australia, said in the announcement.
Dive Insight:
AWS is pitching enterprises on its trio of new frontier agents as CIOs work to reap the benefits of AI.
Most AI agents function within a narrow workflow, said Tom Coshow, senior director analyst at Gartner. AWS distinguishes its frontier agents from other agentic services available because they can be sent off to work with little to no human direction for a significant length of time, according to Coshow.
“An AI agent that can work continuously on a project and work on a project for a long time, especially mostly autonomously, those are pretty rare in what you see being built by enterprises,” Coshow said.
The Kiro autonomous agent handles tasks such as triaging bugs, improving code coverage and completing tasks described by users. The agent can connect to tools including Jira, GitHub and Slack to maintain context as it works.
The second frontier agent in the set, AWS Security Agent, reviews design documents against organizational security requirements and common vulnerabilities, AWS said. Meanwhile, the AWS DevOps Agent provides on-call availability when incidents happen within AWS or across multicloud environments, and is tasked with responding to and finding the root cause of issues.
“If you’re an engineer, you spend a lot of time complaining about the fact that all you do is put out fires, read alerts and never actually have time to build anything,” Coshow said. “If these AI agents can lift that burden, that gives more productivity time.”
It’s not just hyperscalers developing autonomous AI agents. CIOs are also zeroing in on the benefits of AI agents in software development.
Global financial services firm BNY deploys what it calls digital engineers to conduct vulnerability management within the bank’s ecosystem. The autonomous agents fix low complexity code and report more difficult issues to managers, said Leigh-Ann Russell, CIO and global head of engineering at BNY.
The digital engineers free up BNY’s human engineers to complete more high-value projects, Russell said during the Gartner IT Symposium/Xpo in Orlando, Florida, in October.
“If you identify a role in an organization and you take that role and you pick one workflow that is kind of a bottleneck, anytime you can put an AI agent in there, if it will work reliably, it can be a big benefit,” Coshow said.
