Onedot's DIY Onboarding Demo Still Needs a Target Model

The Onedot session is a live walkthrough of AI mapping for DIY retail. Stephan Spijkers and Rafael Foerg showed supplier product data going into a PIM, against ETIM and against custom taxonomies.

The page gives Rafael Foerg two titles. The introduction calls him Technical Lead at onedot. The speaker line calls him Global Head of AI Digitalisation & Strategy, Onedot. The source spelled that second title "Gloal". Both titles stay until the recording confirms which one he used that day.

Speakers:

• Rafael Foerg - Technical Lead at onedot, and, on the speaker line, Global Head of AI Digitalisation & Strategy, Onedot
• Stephan Spijkers - Co-Founder, PIMvendors.com

What the session covered:

• The manual job: complex supplier formats, inconsistent taxonomies, and weeks of manual work before a product is online.
• The demo: a live walkthrough of Onedot's AI mapping, with examples of supplier product data landing in a PIM.
• The taxonomies: ETIM, and custom taxonomies, for retailers and distributors in DIY, home improvement and B2B.
• The advice: ways to shorten onboarding and raise accuracy.

Mapping writes a value into the target you point it at. If ETIM is the target, the class and the attribute have to be chosen before the model runs. If the target is your own taxonomy, someone has to own that list. Mapping into the wrong class stores the wrong record.

Product classification: GS1, ETIM and ECLASS is what an ETIM class asks the file to contain. The product data process puts onboarding before enrichment and publication, which is the sequence this demo sits inside. A later write-up of an Onedot mapping session, AI mapping still needs an owner for the target model, makes that ownership point on its own. AI-ready product data is the bar for the record after mapping, when an assistant is going to read it.