.webp)
I just got back from Americas Procurement Congress (thanks for hosting us Procurement Leaders | A World 50 Group Community), and one conversation kept surfacing in nearly every session I attended: the age-old debate between consolidated platforms and purpose-built point solutions.
It’s a debate worth having. But I think a lot of people are landing on the wrong side of it, and I think the reason why comes down to a misunderstanding of what AI actually does well.
We’ve been here before.
If you’ve been in enterprise technology for any length of time, this argument sounds familiar. When ERPs and CRMs first emerged, the pitch was simple and seductive: one platform, one login, one vendor, one throat to choke. SAP, Oracle, Workday, and others spent the next two decades acquiring company after company trying to deliver on that promise.
And yet, almost nobody actually uses the full integrated stack from any of these vendors. Why? Because the stack isn’t really integrated. It’s a collection of products that were built independently, acquired at different times, and stitched together with varying degrees of care. A Frankenstein’s monster. The user experience reflects that reality. It’s functional, but it’s nobody’s favorite tool for any particular job.
The pitch of consolidation has always been more compelling than the reality of it. It is a pitch that IT almost always finds more compelling than the end users of the tech stack.
Now AI enters the conversation, and the pitch is back.
At APC, I heard versions of the same idea… AI is going to be the solution that finally makes the single-platform dream real. Co-Pilot will sit on top of everything. One agent to handle it all.
I think that’s exactly backwards.
The true promise of AI isn’t consolidation. It’s specialization. The magic of AI is that it can be trained, tuned, and built around very specific domains, very specific data sets, and very specific workflows. That domain expertise is what makes the output valuable. A general purpose AI assistant that knows a little about everything is useful in the same way a Swiss Army knife is useful. It’s a clever idea, it fits in your pocket, and none of the tools on it are particularly good. Even the knife is mediocre.
Note before anyone calls me out: the actual Swiss Army carries a “soldier knife” that is itself a specialized version of the original, with fewer tools and a locking blade.
Professionals don’t use Swiss Army knives. Surgeons don’t grab cheap razor blades off the shelf even though they’re perfectly sharp. Woodworkers pay a premium for Festool even when there are excellent, cheaper power tools available. The nuance of each discipline demands nuanced tooling. Professionals invest in that specificity because the time saved and the quality gained are worth many times the cost of the tool itself.
AI makes specialization more powerful, not less.
Here’s what I think people are missing. AI doesn’t reduce the need for domain-specific solutions. It amplifies the value of them. An LLM is a foundation, not a finished product. The difference between a generic AI assistant and a purpose-built AI tool is the data underneath it, the workflows around it, and the years of domain knowledge baked into how it operates. You can’t just wave a magic wand and have a horizontal platform suddenly understand the intricacies of chemical sourcing, or pharmaceutical procurement, or raw material supplier qualification.
What AI is genuinely great at, though, is integration. Purpose-built tools that are designed around specific functions can now talk to each other, share context, and work together in ways that weren’t possible five years ago. The glue is getting better. That means you can have the best sourcing tool, the best contract management tool, the best spend analytics tool, and AI can help them operate as a cohesive system without anyone having to shove them all into one monolithic platform.
That’s the actual unlock. Not one platform to rule them all. A constellation of sharp, specific tools that AI helps orchestrate.
We are years, if not decades, away from true platform convergence. And frankly, I think the idea itself runs counter to what AI is best at. The further you generalize, the more you lose the depth and specificity that makes these tools transformative in the first place.
So if you’re a procurement leader evaluating technology right now, I’d push back on the consolidation narrative. Ask instead, “does this tool understand my category, my workflows, my data? Was it built for what I actually do every day?”
Because that’s where the real ROI lives. Not in fewer logins, but in better outcomes.
Chase Macaione is the Global VP of Sales at Valdera, an AI-native platform purpose-built for chemical and raw material sourcing.