Mateusz Srodkowski | SOPA Images | Light Rocket | Getty Images
Amazon’s The cloud division on Tuesday unveiled AI-enabled software designed to help customers understand and recover from outages.
DevOps Agent is an Amazon Web Services artificial intelligence tool that predicts the cause of technical problems using input from third-party tools such as: data dog and dyna trace. AWS said customers could sign up to use the tool in preview on Tuesday before Amazon starts charging for the service.
AWS’ AI outage tools are intended to help businesses more quickly identify the cause of outages and implement fixes, Swami Sivasubramanian, vice president of agent AI at AWS, told CNBC. This is what site reliability engineers (SREs) do at many companies that provide online services.
SRE strives to prevent downtime and take immediate action when an actual incident occurs. Startups like Resolve and Traversal have started marketing AI assistants for these professionals. Microsoft’s Azure Cloud Group introduced SRE Agent in May.
Rather than waiting for an on-call staff member to figure out what happened, AWS DevOps Agent automatically assigns work to agents who investigate different hypotheses, Sivasubramanian said.
“By the time an on-call operations team member makes a call, they have created an incident report that includes a preliminary look at possible outcomes and then makes recommendations on what can be remediated,” Sivasubramanian told CNBC ahead of AWS’s Reinvent conference in Las Vegas this week.
Commonwealth Bank of Australia tested the AWS DevOps agent. AWS said in a statement that the software found the root cause of the problem within 15 minutes, which would have taken a seasoned engineer hours.
The tool relies on Amazon’s internal AI models and those from other providers, a spokesperson said.
AWS has been selling software in addition to raw infrastructure for many years. Amazon started leasing server space and storage to developers early on, starting in the mid-2000s. google, microsoft and oracle It continued.
Since the launch of ChatGPT in 2022, these cloud infrastructure providers have sought to demonstrate how generative AI models, often trained in large cloud computing datacenters, can speed up the work of software developers.
Over the summer, Amazon announced Kiro, a so-called vibe coding tool that generates and modifies source code based on a user’s text prompts. In November, Google debuted similar software for individual software developers called Antigravity, and Microsoft sold subscriptions to GitHub Copilot.
Spotlight: Amazon rolls out AI-powered tools to help large AWS customers update outdated software
