AI Onboarding Session: Three Vendors, Three Different Jobs

AI onboarding is three different jobs, and this session put a different vendor on each one. Name the job that is actually slow in your catalog before you sit through another demo.

The speakers, in this order:

• Tony A. Majidi of Akeneo talked about AI on the way product information comes in, with fewer misses in the record.
• Bernhard Bicher of Onedot showed cleaning, structuring and enriching product data as an automated step.
• Alexander Grunewald presented with Dirk Peter Grunewald of Digital ConsultingHaus LLC / GmbH. They showed AI connecting and syncing data from several sources into one product view.

The LinkedIn URL on Alexander's name points at a profile slug for Dirk Peter Grunewald. Both names stay as published. The recording is the check.

The three jobs stay separate. Akeneo was on the system that holds the record. Onedot was on the clean-and-enrich pass. Digital ConsultingHaus was on the pipes between sources. If your week goes to fixing supplier files, the second job is the one to test. If the same attribute arrives from three systems and they disagree, the third job is the one to test. Fluency in the demo tells you which speaker was comfortable. The slow job in your catalog is a separate test.

Onboarding sits inside implementation. The PIM implementation guide spells out the sequence: the model, the supplier feed, then the channel. An AI step on an attribute nobody has defined stores the wrong value faster. Product data quality and governance is about who owns the rule that model is supposed to follow. If the cleanup is meant to let an agent read the record later, AI-ready product data says what "ready" has to mean before you turn a model loose on the file.

The session closed with a discussion of where AI-assisted onboarding goes next. The demos are on the recording. Other sessions from the same channel are on the PIMvendors YouTube channel.