TheChek News Flow

The Takeaway

Amazon Web Services is developing AI to automate functions in its sales, business development and other groups where it recently laid off hundreds of employees or even more.

For instance, AWS has been developing an AI agent that helps its sales employees provide customers with fast answers to technical questions, according to two people who have used it. The agent handles some of the work of thousands of AWS technical specialists—employees well versed in areas like cybersecurity and server networking, said the people. The company laid off numerous technical specialists in the recent job cuts, according to a former AWS employee.

An AWS spokesperson confirmed it is developing such an agent, which “aggregates specialist knowledge from across AWS,” enabling those employees to “focus on the most complex, high-value customer challenges,” the spokesperson said.

AWS last week also launched an AI agent that helps its salespeople coordinate with consultancies and other business partners on cloud customer deals, handling basic tasks like updating customer records after sales meetings. The agent also prompts those partners, ranging from Accenture to Insight Enterprises, to include more information about sales leads so AWS sales teams can determine if the leads are viable or decide which ones to focus on first.

Such business partners interact with more than half the cloud provider’s sales leads and active accounts, according to an AWS spokesperson.

Several current and former AWS sales managers and business development staff say these various automation efforts appear to be aimed in part at handling the work of groups hit hard by layoffs earlier this year.

Having squeezed efficiency from every corner but labor, Amazon sees the sales process as “ripe for automation,” said Jason Strimpel, a managing director at Andersen Consulting who helps large companies use AI.

“There’s so much commodity-like work that goes into researching a company, figuring out who the buyer is, gathering all that data, and putting it into a standardized format that can then be read by the next human being,” he said.

‘Higher-Level Work’

AWS spokespeople say the company isn’t trying to replace workers with AI but rather is seeking to relieve them of repetitive tasks so they can focus on higher-level work, such as helping customers formulate their AI spending plans.

“AI was not the reason behind the vast majority of recent role reductions,” an AWS spokesperson said in a statement. “These changes were about continuing to strengthen our culture and teams by reducing layers, increasing ownership, and helping reduce bureaucracy to drive speed and ownership, and be set up to invent, collaborate, be connected, and deliver the absolute best for customers.”

The spokesperson said the company continues to hire people to the affected teams but declined to say how many people are currently working at AWS, something Amazon doesn’t disclose.

The Information previously reported that AWS’ employee base swelled from approximately 26,000 at the end of 2017 to around 140,000 in late 2022 and then fell 8% or so to 115,000 by the end of 2023, following layoffs around that time. AWS has periodically laid off staff since then, including during a period between October 2025 and January this year, when Amazon overall cut 30,000 roles.

Amazon is one of many large tech companies asking employees to use more AI while it cuts staff, especially as they are using much of their cash reserves to invest in data centers, crimping free cash flows. After Microsoft laid off employees in its sales group last year, for instance, executives told remaining employees to use AI agents to manage the additional customer accounts they were handling. And Amazon CEO Andy Jassy, in a companywide memo last June, said he expects the company’s internal usage of AI to result in a reduction of its corporate workforce over time.

Other Amazon leaders are reportedly planning to use robotic automation to do the work of hundreds of thousands of humans by 2033.

Such cuts are contributing to broader worries about AI’s impact on labor, especially as the leading AI firms have said white-collar work will face widespread disruption in the coming years.

Automating Technical Expertise

Amazon is working on an agent that automates part of the role of partner solution architects, which help software providers and consultants develop applications and run computing jobs. AWS has hundreds of solution architect employees.

AI agents that automate multistep tasks are still in their infancy in much of the corporate world, of course, but corporate spending on Anthropic AI coding and other agents has surged. That has set off a mad dash among other sellers of enterprise apps and AI, including AWS, to provide similar tools for creating and managing agents that handle complex coding projects or that can use existing enterprise applications the way humans do but faster.

For AWS and other AI providers, getting employees who interact with customers to use internal AI agents can make it easier for them to sell such products to external customers, who may be initially skeptical of agents and their high costs.

Outside sales, AWS is looking for ways AI can automate roles involving technical expertise. That includes AI coding tools that help engineers handle the company’s codebase and develop new features. Beyond engineering, Amazon is working on an agent that automates part of the role of partner solution architects, which help software providers and consultants develop applications and run computing jobs on AWS, according to two people involved in the discussions about the agent effort. AWS has hundreds of solution architect employees.

AWS has considered using the partner solution architect agent to handle inquiries from new partners who sign up to work with the cloud provider, said a person who was briefed on the matter. These partners will have the option to get additional help from humans, the person said.

When asked about the partner assistance agent in development, an AWS spokesperson said the company is “continuing to develop agents to support specific tasks or processes in the partner journey. These agents are available to all partners, allowing our partners to receive faster feedback and approvals to drive their business forward.”

AI for Sales Staff

The AI agent that helps Amazon’s salespeople coordinate with consultancies and other business partners can also sift through sales leads partners have uploaded to its customer management system. It then asks the partners for more details to determine if those leads need immediate attention, said Matt Yanchyshyn, an AWS vice president.

While partners submitted “millions” of leads in 2025, AWS sales teams haven’t even looked at “hundreds of thousands” of them because there wasn’t enough data to assess their viability, he added.

In terms of revenue, the AWS Marketplace has grown into a “multiple tens of billions dollar business,” and using AI helps AWS sellers keep up with the sheer volume of sales leads, he said.

Long before the emergence of AI agents, AWS found ways to use AI to automate various kinds of sales and partner management functions.

In 2024, AWS launched an AI tool software providers such as Snowflake, CrowdStrike and Datadog could use to evaluate their applications for technical glitches and security vulnerabilities before launching them in AWS’ app store. Previously, human partner solution architects handled the vetting.

That same year, AWS also quietly launched an AI tool that helped its salespeople identify software partners whose products or expertise could assist them in closing new cloud customer deals. Human partner sales managers previously handled this task, according to current and former AWS employees.

Yanchyshyn said using AI to match partners with customers is a slam-dunk move for accelerating the sales process.