AI agents will be biggest technology change ‘since the beginning of cloud computing’

Amazon’s (AMZN) AWS is getting a boost with a line of what the company refers to as “frontier agents,” AI software that Amazon said can work autonomously, perform multiple tasks simultaneously and across several agents, and do so independently.

The agents include Kiro, an autonomous agent for software development; AWS Security Agent, which is meant to ensure that business applications are secure as they’re being built; and AWS DevOps Agent, which is meant to help ensure those applications and services stay online or, when they go down, are up and running faster.

The company also announced its Tranium 3 chip is up and running on its AWS servers.

“The agentic era is arguably going to be one of the biggest technology transformations since the beginning of cloud computing,” explained Swami Sivasubramanian, the company’s vice president for agentic AI at AWS.

“It goes beyond where traditional LLMs [large language models] are,” he said. “It is actually going to move from … just answering what is in a chat bot to actually doing things on our behalf, and that’s why we are really excited about this area.”

Amazon isn’t the only company working on AI agents. Google (GOOG, GOOGL), Microsoft (MSFT), and OpenAI (OPAI.PVT), among others, are developing or actively deploying AI agents. The goal is to turn AI chatbots into truly useful applications that can perform tasks for users with minimal intervention.

Amazon has reported early successes with its own AI agents. The company said it saved 4,500 developer years and $250 million in capital expenditures using an agent it developed for its internal software development upgrades.

And the trio of new agents is no different.

“Internally, we are seeing even our own Bedrock team was able to rebuild its inference platform with [a] really small team, what could have been a year, [in] less than actually [a] few months with … six to eight people,” Sivasubramanian said.

Amazon’s Bedrock is the company’s AI platform. It provides customers with access to a slew of AI models, ranging from Amazon’s own Nova models to OpenAI’s GPT, Meta’s (META) Llama, and Anthropic’s (ANTH.PVT) Claude.

Amazon’s new Kiro, Sivasubramanian said, is designed to help developers put together apps, acting as a team member that can pick up tasks, begin working on them, and send out code for review.

AWS DevOps Agent is designed to function as a kind of emergency worker. If an application or service goes offline, it can begin attempting to figure out the problem and provide potential solutions for the outage.

That could save workers a lot of time they would normally spend trying to identify the cause of an error or devise a fix.

Security Agent, meanwhile, engages with developers building applications from the design phase to code review and, finally, to penetrative software testing, Sivasubramanian said.

Amazon is working to ensure it stays in the top tier of the broader AI conversation. After seemingly falling behind OpenAI, Google, and Microsoft, the company has since made a series of announcements, including working with OpenAI, that have helped boost its standing in the industry. Now it has to hold on to that position.

Email Daniel Howley at dhowley@yahoofinance.com. Follow him on Twitter at @DanielHowley.

For the latest earnings reports and analysis, earnings whispers and expectations, and company earnings news, click here

Read the latest financial and business news from Yahoo Finance.