#130: How Product Data Management Was Born — And What It Means for AI Today

The man who helped name an industry sits down with Stay Sharp in Digital Engineering to trace the winding path from Product Data Management's humble origins to today's AI ambitions — and delivers a blunt warning: fix your data integration first, or AI will fail you.About This EpisodeBrion Carroll, CEO/Principal Consultant at Digital Solution Group, is not just a PLM practitioner — he's one of the people who built the very first PDM system and coined the category name. In this episode, Brion joins Juliann Grant and Jonathan Scott for a rare firsthand account of how product data management came to be, how it evolved into PLM, and why decades later, the industry is still wrestling with the same fundamental challenge: getting data to flow across systems.Whether you're deep in a PLM implementation or just starting to think about your AI strategy, this episode is a grounding, practical reality check you won't want to miss.Key TakeawaysPDM was born as a defensive move — Computer Vision built it to protect its CAD customer base from competitors, not to transform the industryThe first PDM system sold for $1 million per seat and handled only four functions: backup/recovery, revision control, access/security, and archive/restorePortability was a watershed moment — Brion famously declared from a conference stage that their product would run equivalently across three competing operating systems, and delivered on it in six monthsThe evolution from PDM to PLM wasn't a grand vision — it was vendors needing new features to sell, starting with workflow engines to manage product lifecycle stagesWeb technologies (Java, JavaScript, TCP/IP) democratized access and pushed the shift toward client-server and eventually true cloud architecturesToday's AI problem is yesterday's integration problem — AI applied inside a silo gives narrow insight; real business intelligence requires harmonized data flowing across all systems~4% of AI initiatives have yielded measurable value (per BCG research cited in the episode), and Brion argues dirty, siloed data is the reason whyWhat We DiscussBrion takes us back to the mid-1980s and Computer Vision's "Project David" — the small team tasked with building a data management system to protect the company's installed CAD base. From there, the conversation moves through the birth of the PDM category name, the wild competitive landscape of early vendors (Sherpa, Workgroup Technology, CDC), and the technical leap of making software portable across VMS, Unix, and IBM VM/CMS environments.We then trace the shift to PLM — why workflow management changed everything, how web technology opened the door to broader connectivity, and how the retail industry became an unlikely early adopter of enterprise PLM. The conversation closes with a clear-eyed look at today's AI landscape and why Brion believes companies that skip the hard work of data integration will find AI disappointing at best and damaging at worst.Guest ResourcesBrion Carroll— CEO & Principal Consultant, Digital Solution GroupWebsite:digitalsolutiongroup.netEnjoyed This Episode?Leave a review, drop a comment, or share this episode. Reach out anytime at [email protected].👉 Subscribe and never miss an episode — and don't forget to Stay SharpMusic is considered “royalty-free” and discovered on Story Blocks.Technical Podcast Support by Jon Keur at Wayfare Recording Co.© 2026 Razorleaf Corp. All Rights Reserved.