Circana hires Ode to build AI workflows, with experts handling uncertain work

Circana wants AI to speed up market intelligence without sacrificing expert oversight. On October 29, 2026, it announced an engagement with Ode with Anthropic to build workflows for data processing, trend forecasting and product classification. The announcement describes a development effort, not a finished product or demonstrated gains in speed and quality.

The work targets how Circana produces and delivers information to clients. The company serves consumer packaged goods businesses, durable-goods manufacturers and retailers, using data and predictive analysis to help them measure market share and understand consumer behavior. Its stated goal for the engagement is faster, more flexible insights that clients can use to make decisions.

A shared platform, with an expert backstop

Circana and Ode are building a common platform where teams can create, test and deploy AI agents. The platform’s stated controls include measuring every agent against cases verified by experts. That puts a reference set of checked work into the development process, rather than describing deployment alone as the goal.

Uncertain work is to go to Circana specialists. The design therefore gives people a specific role inside the workflow: handling work that needs expert attention, not simply reviewing the overall AI initiative. Circana says the end-to-end workflows will be guided by experts across three initial areas:

• Data processing: one of the first targets for expert-guided AI workflows.
• Trend forecasting: applying the initiative to work focused on anticipating market changes.
• Product classification: another initial focus, alongside processing and forecasting.

The companies describe the intended output as production-ready capabilities on a reusable foundation, rather than standalone productivity tools. Their ambition is to support additional client solutions over time. That makes the platform a proposed base for ongoing development, not just a collection of separate tools for the first three tasks.

“AI works best when it gives experts more room to use their judgment,” said Chris Taylor, CEO of Ode. “Circana’s analysts carry instincts that took years to sharpen. We’re thrilled to work alongside them and amplify the impact of their expertise.”

Chris Taylor, CEO of Ode

Outside engineering, internal responsibility

The engagement also includes building an in-house agent team. Ode is training Circana engineers to build and deploy agents, while Circana operators will own the goals and quality of agents in their areas. Technical development and operational responsibility are both part of the arrangement; the announcement does not describe a simple handoff to an outside supplier.

Ode’s role is implementation: it combines Anthropic’s AI models with engineering teams that help organizations identify useful applications and build systems around them. For Circana, the announced scope reaches beyond access to a model. It includes the platform, workflows and training needed to put agents into the company’s own processes.

Michael Rosenthal, Circana’s president of operations, says removing manual steps and sending exceptions to experts should make work faster and more consistent without compromising quality or oversight. Those are the intended benefits, not published measurements. Circana says it will share results and case studies as the work scales; the announcement provides no quantified speed or accuracy gains.