Dataiku has launched Dataiku Headless, a new capability that integrates its natural-language agentic interface, Dataiku Cobuild, into popular coding tools to help technical teams build, manage, and modernize data workflows.
Unveiled at the company’s annual Succeed conference, Dataiku Headless enables developers to work within coding environments such as Claude Code and Codex while using Cobuild to create and modify Dataiku projects.
The capability allows coding agents to turn developer instructions into visual data pipelines, datasets, transformations, machine learning models, dashboards, and AI agents within the Dataiku platform. These assets can then be reviewed, managed, and deployed by other members of the organization.
Dataiku Headless also supports legacy workflow modernization. Coding agents can analyze existing workflows, including Excel workbooks and Alteryx projects, and rebuild them as governed, visual Dataiku projects. This approach can help organizations accelerate large-scale migration initiatives and reduce manual development work.
For platform administrators, the capability can also investigate failed jobs across multiple projects, help identify operational issues, and assist with remediation through project and administrative setting changes.
According to Dataiku, the technology addresses the challenge of making AI-generated code and workflows accessible, reviewable, and manageable across business teams rather than leaving them confined to individual developers.
Dataiku Headless is available through the Claude Code and Codex marketplaces. It is also compatible with Cursor, Snowflake CoCo, GitHub Copilot, AWS Kiro, and other agentic coding tools.
More information is available at https://www.dataiku.com/blog/dataiku-headless.
