Centric’s New Retail AI Targets Product Paperwork

Centric Software is bringing a new AI platform to the work behind retail products: assembling specifications, finding product information and coordinating decisions before goods reach a shelf. Announced on September 10, Centric AI combines a conversational assistant, connections to outside AI systems, task-running agents and specialist applications. The company plans to demonstrate it at the National Retail Federation’s European show in Paris, September 15–17.

For a clothing brand, the promise is more specific than generating a picture of a jacket. A design has to become information that other teams can use to develop, source and make the product. Centric is pitching AI inside that process. Its public materials describe what the platform is intended to do, but do not establish that every advertised function is available to every customer on the same terms.

From a sketch to a parts list

One example is the BOM & Tech Pack Agent. BOM means bill of materials: the list of components needed to make a product. Centric says the agent can take a sketch or supplier document, find the closest existing style and assemble details covering fabric, trim and packaging. That targets the translation from an idea into structured production information, rather than the visual idea alone.

The company’s product page adds an important qualification. It says agents flag details they cannot confirm, such as an assumed quantity or an unverified supplier price, for a person to resolve. That is a claimed behavior, not a result tested for this article. In a hypothetical jacket workflow, a neatly completed row of fabric information would be useful only if the team could tell whether it came from a confirmed specification or still needed a decision.

Centric AI Companion is the conversational part of the platform, intended to retrieve product and commercial information and trigger work inside existing tools. Centric MCP connects external AI systems with Centric’s data and permissions. Centric AI Studio handles product ideas, design and visual content. These are different functions under a common platform name; access to one should not be taken as evidence of access to all the others.

Big savings claims need their boundaries

Centric reports that customers using its AI alongside its broader software reclaim 90–98% of time spent on selected manual work. Its product page explicitly says the results come from the combination of AI, its software and human judgment. The inspected materials do not provide enough detail about customer samples, measurement periods or comparisons to treat that range as a forecast for another retailer. It is not evidence that almost an entire employee’s workload disappears.

There is a wider operational reason to pay attention. In an April overview of retail technology, the National Retail Federation described businesses trying to connect existing systems and scale AI beyond pilots. It identified inconsistent data, differing processes and unclear ownership as obstacles. That trade-group account provides context for Centric’s pitch, rather than an independent validation of the new platform. The organization also runs the show where Centric will demonstrate it.

The difference becomes concrete when information moves between teams. A hypothetical supplier document might describe a material differently from the retailer’s internal record. Filling a form quickly would solve only part of that problem. Establishing which record is authoritative, and showing unresolved details to the right person, would determine whether the output could move the product forward. Centric’s emphasis on existing product data and human exceptions addresses that need in its stated design; the announcement alone cannot show how reliably it works.

What is available to inspect

The public product page directs interested teams to request a demonstration. Neither that page nor the announcement provides a public price list or a feature-by-feature rollout schedule. Those are limits of the published information, not proof that the software is unavailable. The immediate development is a more unified AI offering around Centric’s retail systems. The unanswered question is how its individual capabilities behave with a customer’s actual records, permissions and unresolved decisions.