We built one of the best procurement platforms on the market. <span style="color: #0D5C6E;">Then we rebuilt it.</span>

Three years ago we set out to build a procurement platform with AI at the core. It worked. And then, before the product was even three years old, we tore the whole thing down and rebuilt it.

The question we get asked more than any other is what “AI-native” actually means. It is on just about every company website now, so it is a fair thing to be skeptical about. Our answer happens to be the same as our reason for rebuilding.

It means everything is a moving target based on how fast things have evolved. That is the part that has made building Levelpath the most fun either of us has had in this industry.

When we started, ChatGPT had not even had its moment yet. Since then the pace has been absurd. Claude Code. Cursor. Codex. AI agents holding down real jobs. Thousands of individual breakthroughs, most of them arriving faster than anyone could absorb them. There is no way to bolt that much change onto an existing platform, not even one that has not had its third birthday.

That is the lesson we keep relearning: with AI, you need a new platform to keep up with this type of change. That is why we can't wait to introduce you to Ranger, Levelpath's Autonomous Source-to-Pay platform.

‍

So why rebuild?

Levelpath V1 was already one of the most AI-native platforms in procurement, running in some of the largest companies in the world: Ace, Western Union, Amgen, New York Life. Procurement and AI analysts at Gartner and IDC put us at the top. Our customers loved us. Most companies would ride that for years.

But every customer we have is trying to do the same thing: drastically increase throughput and coverage without adding headcount. They are all leaning on AI to push as much of source to pay as possible into running itself. Our three-year-old V1 stack, a baby in enterprise SaaS terms and ancient in this one, could not take them there.

Meet Ranger

Ranger is the next generation of Levelpath, and the name matters.

Think about what a park ranger actually does. They don't walk every trail and inspect every tree one at a time. They know the whole landscape. They see what's coming, get ahead of it, and spend their energy on the calls that actually protect the place. They direct the work instead of drowning in it.

That's the shift Ranger makes. The work no longer waits for a person to push it forward. A request for new software comes in, and Ranger routes it, pulls the supplier and contract data from the systems you already run, checks it against your policy, and moves it toward a decision. It handles what it can on its own, and it knows when to bring in a person, the way a good teammate does. The copying, the chasing, the endless status checking quietly disappears. Your team stops processing the work and starts directing it.

‍

What we are automating, and what we are not

We should be precise here, because the industry has not been.

You cannot automate procurement. Procurement is negotiation and judgment. It is the supplier relationship that took four years to build, the trade-off with no clean answer, the risk call somebody has to own. That is the business of procurement, and it will never be autonomous. It shouldn't be.

What you can automate is source to pay. The routing. The copying. The chasing. The policy checks, the status updates, and the three decades of accumulated machinery that landed on your plate because software could record the work but could never do it. That is what Ranger takes on, all of it, running quietly in the background with your policy applied and a full audit trail on everything an agent touches. Real money and real risk mean trust has to be built into every step, from governance and security on down.

Take the machinery off the plate and what's left is the job people actually signed up for. That's the trade. Ranger doesn't shrink a procurement team. It hands one back the business of procurement.

Now is the moment. It is AI-native or bust, and honestly, AI-native might not even be enough. Unlike any other era in the history of software, the reinvention journey never finishes. It just keeps moving.

Over the next decade, we believe the entire multi-trillion-dollar function of procurement will get rewritten. We rebuilt Levelpath because we intend to help our customers write it.

We've spent more than a decade in this industry and the best part of it has always been the people: the practitioners who welcomed us, challenged us, and kept pushing us to be better.

We remember that first spark of what AI could do. We've spent every day since chasing it into procurement. The old era recorded the work. This one does it.

We have never been more excited to be building software.

Let's go.

‍