Dataiku Launches Headless to Connect Claude Code and Codex to Its Governed AI Platform

Headless brings Cobuild, Dataiku's natural-language agent interface, into the environments where developers already work. From those tools, developers can create and change data pipelines, models, dashboards and custom AI agents. Instead of producing standalone scripts, the work is saved in Dataiku as visual workflows, datasets, transformations and models that colleagues can inspect, maintain and govern.

The Problem It Targets

The launch addresses a growing concern for companies adopting coding agents. The tools can produce software quickly, but the code they generate can be hard for anyone other than its author to review or run in production. Dataiku describes Headless as a control layer between coding agents and its platform. Agents can inspect and operate Dataiku and delegate building work to Cobuild, so the output stays inside a governed system.

Jed Dougherty, senior vice president of AI and platform at Dataiku, said developers "aren't going to stop working in Claude Code or Codex, and we don't think they should have to." He said the product "hands back a Dataiku project instead of a script only they can read."

Migration and Administration

Dataiku is also positioning Headless as a migration tool. Developers can have a coding agent analyse existing Excel workbooks or Alteryx workflows and then tell Cobuild to rebuild them as Dataiku projects. The company said agents can repeat this across hundreds or thousands of workflows, which could speed up modernisation programmes that would otherwise need a lot of manual rework.

Administrators can use coding agents to find failed jobs across several projects, diagnose problems and change project configurations or administrative settings.

Availability

Headless is available as a plugin in the Claude Code and Codex marketplaces. Dataiku said it also works with Cursor, Snowflake CoCo, GitHub Copilot, AWS Kiro and other agentic coding tools. Pricing was not given in the announcement.

Why It Matters

Data and analytics vendors are trying to stay relevant as more development moves into general-purpose coding agents. Rather than asking engineers to use another interface, Dataiku is putting its platform behind the agents they already use and offering governance and visibility as its value. Banks, energy companies and government entities in the Gulf that are expanding their use of AI tend to face strict audit and compliance requirements. For them, knowing who built a pipeline, how and on what data counts for as much as how quickly it was built.