If you’ve ever done a service catalog migration, you’ll know exactly what I’m talking about. Someone hands you hundreds of spreadsheets, or maybe some old Visio diagrams from years ago.

You have a few kickoff calls. The scope looks “fine.” Timelines feel tight, but doable. And you think, yeah, we can knock this out in four months.

Then four turns into five, and before you know it, six months have gone by.

Before I joined Echelon, a customer of mine was moving off a legacy ITSM tool onto ServiceNow. The goal was to migrate around 200 catalog items. We had four full-time developers. The plan looked solid, and a four-month timeline felt realistic.

But things didn’t go as expected. It wasn’t because the team lacked experience. 

They were strong. The delays came down to two main issues:

1. Technical challenges we hadn’t anticipated
2. Bottlenecks across all the people involved

It doesn’t have to be that way, though.

In this blog post, I’ll share how we, as a ServiceNow partner, used an AI developer to cut a six-month migration project down to just six weeks.

Complexity often hits after you start building

On paper, building a catalog item always looks simple. But what starts out small often turns into weeks of effort once the actual work begins. As you get deeper into the build, things get tricky, leading to a lot of back-and-forth.

For instance, what looked like a basic request form turns out to be five requests stuffed into one.

We had a number of catalog items with 50+ variables, 10 or more UI policies, and all of them connected. Update one field, and something else would break. Debugging took forever. 

And testing? Sometimes, QA cycles dragged on longer than the actual development time.

During process and discovery workshops, we tried to simplify. We broke things apart where we could. But even after all the optimization, modernization, and simplification efforts, some items were still heavy.

They were “complex” for a reason. The business needed them that way. There’s always a balance to strike between technical best practices and user experience.

Then came the bottlenecks caused by the people involved.

The back-and-forth slows everything down

Here’s how the process looked. The business gave their requirements to a business analyst. The analyst would validate, clarify, suggest, and re-document, while getting the process owner’s sign-off with each step. Once signed off, the requirement made its way to the developer.

And that’s when the questions started. The developer needed more clarification.

So it went back to the analyst, who had to follow up with the process owner. Sometimes that person wasn’t available. Other times, they needed more time to think it through. If the dev team was offshore, even the smallest clarification meant waiting another day.

This back-and-forth didn’t just slow us down, it stretched out timelines without anyone realizing it. One question here, one delay there, and suddenly you’re weeks behind.

While you can add more people into the mix — such as more analysts, more calls, or more standups — it doesn’t solve the problem. You just end up adding overhead, and still, everyone is just waiting on someone else. It’s not about bad people or bad processes. It’s just how work tends to flow. And that flow is slow.

When you’re trying to move fast or scale delivery, this kind of lag becomes your biggest blocker. Most teams don’t even see it until it’s already baked into the project timeline. But once you spot it, you realize just how much time is being lost.

The fundamental shift with AI

Knowing how this dated, traditional model worked, and then transitioning into Echelon, it fundamentally changed and accelerated the entire process. With a similar type of customer who had even more catalogs to migrate, here’s how Echelon’s AI developer changed the game for us:

Instead of routing every piece of input through five people, the business process owner directly uploaded their requirements and documents.

Then the AI developer analyzes it and asks follow-up questions like:

“I see a process flow with 3 branches, but only 2 triggers. Should there be a 3rd?” 
“This step mentions a group approval. I can’t find a group with this name in the [sys_user_group] table. Would you like me to create one?”

The kinds of things a seasoned developer would ask. With AI, these questions came instantly. The process owner could immediately respond, without having to wait for the analysis, which typically took until the next day. So that meant there was no need for bouncing back and forth.

And once everything was clear, the AI proposed the plan to build the catalog item, and once approved, the catalog item was ready to test with the ATFs within minutes.

That kind of real-time interaction and development velocity changes everything.

You remove the human-to-human lag entirely. While you still have people validating and governing, AI does all the heavy lifting. It reads the input, asks smart questions, fills in the gaps, and builds the catalog item based on the acceptance criteria defined by you.

This shift unlocks infinite capacity for ServiceNow partners and platform owners alike, reducing delays and letting information flow easily from the source to the solution rapidly.

Initially, I thought the time savings would come only from the technical development phase.

While a task that usually took a developer a few days now happens in hours, the biggest shift was that we could save a lot of time even before the build began.

The time that used to go into clarifying requirements, chasing approvals, and getting alignment is now reduced by more than half. That was something I didn’t expect at first — but once I saw it, it made total sense.

That’s how we cut a 6-month migration down to 6 weeks. We didn’t cut corners. We didn’t skip steps. We just changed the way we worked.

What this means for partners or a platform owners

If you’re leading a ServiceNow team, this probably hits close to home. You’ve got a roadmap. You’ve got real business needs. You’ve got smart people on your team. But the thing that slows everything down is the pace of delivery.

Your backlog grows. Your best ideas sit in a PowerPoint somewhere, waiting to get scoped. And even when the team starts, the smallest thing can stall progress.

That’s where an AI developer changes the equation.

You don’t need perfect documentation. Just give it what you have—a process doc, a flow, or even a few lines of description. Then the AI asks you smart follow-up questions, fills in the gaps, builds it, and gets it ready for your review.

You’re no longer stuck waiting on 3 layers of translation just to get started.

Now your roadmap feels doable again. You stop shelving ideas that felt too complex to tackle. You start delivering business outcomes without having to add to your team.

The catalog item that used to take four weeks is now done in a day — with documentation and ATFs automatically created.

Your developers aren’t stuck re-explaining stories. They’re reviewing what’s built and moving on to the next one. Your team becomes a delivery engine. 

And instead of asking how to keep up, you get to ask: What’s possible next?

Summing up…

Service catalog migrations used to be slow because of too many handoffs and too much back-and-forth. That was just the way it worked. But it doesn’t have to be that way anymore.

With the help of an AI developer like Echelon, you can remove the friction entirely. No more waiting days for answers. No more meetings to clarify basic decisions. This isn’t hype — it’s how things work today.

And it happens without burning out your team or blowing up your budget. Once you start building this way, you’ll wonder why you didn’t make the switch sooner.