Kantata launches Agent Studio for building agents in plain language
Kantata has made Agent Studio available for professional-services firms to create and publish custom agents through conversation, with automatic structure checks before publication.
Kantata launched Agent Studio on September 22, making a conversational agent-building tool immediately available to professional-services firms. The company said customers were already using it.
Agent Studio is part of the Kantata Expertise Engine. A user describes a job and delivery standards in plain language, and Kantata’s existing Expertise Agent builds and refines a custom agent through dialogue. Kantata said practice leads, delivery directors and project management office owners can use the process without a developer or AI specialist. Published agents can be started by a person or automatically by a workflow with a single command.
Kantata said it automatically checks every build and edit before publication to ensure inputs and outputs are structured correctly. The announcement did not explain the validation method, evaluation criteria, approval controls, audit logs, model stack or data-handling safeguards, leaving the scope and performance of those checks unverified.
The company also said firms can record a recurring deliverable’s format, escalation rules, level of detail and required outputs once, then apply those standards repeatedly. Kantata claims this makes experienced practitioners’ judgment available across projects and produces more consistent work. The supplied evidence contains no independent measurements of consistency, time savings or customer outcomes.
Kantata cited Basware as a customer example. Basware Chief Customer Officer Mark Johnston said the company wants professional-services staff focused on customers instead of documentation, resourcing and billable-time entry, and that Kantata lets it create agents to handle that work within its processes. The announcement did not identify a specific Basware deployment of Agent Studio or quantify product-specific results.
More news

xAI launches Team Bots for shared workflows

AWS adds xAI’s Grok 4.7 to Amazon Bedrock

OpenAI adds $5 million and up to $5 million in credits to Lenfest AI program
