15 May AI Tools Session: Match the Tool to the Job

On 15 May 2025 the session did not crown an AI tool. It showed three different jobs, and a buyer who treats them as one purchase will fund the wrong one.

The recording:

Speakers:

• Stephan Spijkers, Co-Founder, PIMvendors.com
• Morten Næss, VP of Product, Bluestone PIM
• Hendrik van Zwol, AI Consultant and Speaker

The session showed three tools and did not rank them. Nodemation was shown automating workflows with agents. Fozzles was shown on product content and SEO copy. Bluestone PIM was the example of enrichment built into the PIM, rather than a tool standing beside it. The speakers' line was that a team can try these now, including people who do not write code, on work they already do: cleaning, categorising, product pages, campaign text.

Onboarding is still the hard part. A model that writes a page from a thin record publishes the thinness. The job is to match the tool to the catalog you have.

• Workflow agents (Nodemation in this session) move files and steps. They do not decide your attribute model.
• Content tools (Fozzles here) draft copy. Someone still signs the claim on the page.
• Enrichment inside the PIM (Bluestone PIM here) stays next to the record. Ask what it may overwrite.

Where those jobs sit inside a PIM, and where they belong in a tool next to it, is the split in what AI inside a PIM actually does. The rules for what may be overwritten are in product data quality and governance. Enrichment, workflow and publishing are also a capability question, covered in PIM features and capabilities. Anything you hope an agent will answer later still depends on the record, which is the requirement in AI-ready product data.