AWS Quick Flows now builds automations from a chat
AWS updated Flows in AWS Quick so users build, run and refine automations by describing a process in chat. New deterministic steps and validations check each step before the flow continues.
By JetherVerse News

Amazon Web Services announced on November 20, 2026 an update to Flows in AWS Quick, its AI workspace for business users. Users now build, run and refine automations for themselves and their teams entirely through conversation. A user describes a process in chat, and Quick builds the flow with no coding.
Flows now orchestrate agents, skills and tools, with access to structured data, knowledge and apps. A flow pulls answers from business data a team already reports on, produces ready-to-share documents, spreadsheets and presentations, and saves the results to a team space. Flows also support longer-running workflows and larger volumes of data.
New deterministic steps keep results consistent. Validations check the output of each step against rules the user sets, such as requiring a summary table in a report, before the flow continues. AWS gives an accounts payable example: a team builds a flow in which agents read invoices and match them to purchase orders, while deterministic steps check totals and catch duplicates.
Users run flows wherever they work, on desktop, web and mobile. In the desktop app, flows work with local files and finish tasks in the browser. In the mobile app, users start flows and check progress on the go. The updates are available in all AWS Regions where AWS Quick is available. AWS directs readers to the Flows section of the AWS Quick User Guide and the product page.
Why this matters: the mix of agents for judgment and deterministic steps for checks addresses a known weakness of AI automation, which is inconsistent output. Rules placed between steps catch errors before they reach the next stage, and a finance team gets repeatable results rather than a different answer on each run. The chat interface also lowers the skill needed to start, so non-developers build their own tools.
If you plan to automate a process with AI agents, write down the rules for a correct result first. Then place a validation after each step to check them.
Read the full report at Amazon Web Services.
Source: Amazon Web Services
Get the news in your inbox
The tech stories that matter from JetherVerse News, with the source on every one.

