How Yext scaled AI adoption across every team

Why it matters

Piloting AI is easy. Getting it used across a company is the hard part, and it tends to stall in the same place. A central team becomes the only group that can build anything, and everyone else waits in line.

Yext set out to remove that constraint. The goal was to put automation and AI tools in the hands of every function, from finance to sales to HR, so a team could solve its own problem without filing a ticket and waiting days for someone else to get to it. That meant a platform people across the company could use directly, rather than one gated behind a small group of specialists.

Adoption followed. Monthly automated tasks grew roughly tenfold, and the same foundation now connects those teams to MCP servers and AI tools as they arrive across the business.

→ Explore Tray’s intelligent iPaaS

In their words

What drove the switch from a legacy iPaaS to Tray?

The legacy iPaaS platform was performing its job. We were able to run our data integrations. But we had this roadmap back then to look at tools in order to support our roadmap, and we were focusing on democratization of tools across the company. The whole industry is looking at AI hype and laying a foundation so they can be on that train to leverage the AI tools. We understood that if we have to move teams towards adopting AI quickly and in a faster way, we need a robust iPaaS platform without any limitations which everybody can use and automate their workflows. Tray we felt possessed those capabilities, and we made a switch from the legacy platform to Tray.

When we launched Tray, we were running like 3 million tasks per month. Now we are at 30 million tasks per month. The adoption grew exponentially, which was our goal. What it also unlocked for us is things we are discussing now about leveraging MCPs, connecting to different SaaS platforms. With earlier legacy platforms it would have happened, but not in a seamless way that we are seeing now, where every user is able to take the benefit of connecting to these AI tools. And Tray is that connecting glue between all these platforms and the AI initiatives we are undertaking at our company.

"Tray is that connecting glue between all these platforms and the AI initiatives we are undertaking at our company."

What needs to be true for AI adoption to actually work?

The users have to unlearn something. We have to help them unlearn those skills they have developed over many years working on a specific tool, and try to do the same stuff differently or in a much better fashion, whether adopting some of the boards that are available or developing intelligent workflows. Identify those opportunities, use the AI tools to automate them, and try to identify problems that we haven’t solved till now.

In my role, what we are trying to do is making those tools available. Instead of asking them to just go use AI, we are developing playbooks and boards in action, so they can look at them and see the benefits. Earlier, if you wanted to get a report, you would have to log a ticket and wait for a couple of days. Now there is a board, you can ask a question, and it can generate the report in a few seconds.

As an IT leader, we have to train them, enable them, and introduce the capabilities in a much simpler fashion where they can adapt and use it. That’s when they can see the benefit, and you can make champions in different teams who then act as a catalyst to increase adoption across the organization.

"You can make champions in different teams who then act as a catalyst to increase adoption across the organization."