The data analytics firm is betting that enterprises don’t need faster AI — they need AI that runs on logic they already trust.
Alteryx has unveiled a set of new capabilities designed to move enterprise AI from experimentation into day-to-day operations, anchoring AI agents in the business logic and data workflows that organizations have already tested and approved.
The centerpiece of the release is Agent Studio, a tool that allows teams to package existing datasets and business logic into reusable agents within the Alteryx One platform. Alongside it, a new MCP Server extends those agents into external enterprise applications — including Slack, Microsoft Teams and AI models such as Claude and OpenAI — without requiring organizations to rebuild their underlying workflows.
The approach is a direct response to a persistent concern about enterprise AI: that models generate plausible-sounding outputs untethered from how a specific business actually operates. By running agents on top of workflows analysts have already validated, Alteryx argues the result is AI that is auditable, repeatable and consistent across tools and channels.
The release also includes updates to the Alteryx One desktop application, which now serves as a unified entry point for the company’s Designer tool, cloud services and AI features. A Live Query integration for Google BigQuery allows users to process unstructured data directly where it lives, without moving it or writing code.
On the infrastructure side, Alteryx is expanding deployment options to give organizations more control over where workflows run. Workspace Execution, now generally available, runs workflows in the cloud to reduce dependence on local machines. A Data Bridge feature provides secure access to on-premises data from cloud workflows. Server Execution, coming soon, will allow analysts to manage and schedule server-based workflows from a cloud interface while keeping them running on-premises.
Also Read: 100 Things Google Announced at I/O 2026
The company says its customers ran more than 380 million workflows across environments last year. The new governance layer within Alteryx One automatically versions workflows, assigns ownership and routes them through approval processes — a system aimed at IT teams trying to maintain oversight as AI moves into production.
“AI is only as good as the business logic underneath it,” said Ben Canning, Alteryx’s chief product officer. “Alteryx turns the workflows your analysts already trust into the layer agents run on — so AI stops generating fast guesses and starts doing the work the same way every time.”
