Serving customers in more than 100 countries, in dozens of languages, and using a complex tech stack that's been shaped by years of acquisitions, Lightspeed’s support environment is anything but simple. But instead of slowing them down, that complexity became the proving ground for one of the boldest AI transformations in enterprise support.

In early 2023, Lightspeed became one of the earliest adopters of Fin, Intercom’s AI Agent for customer service, and they’ve been scaling it ever since.

What started as a beta experiment has become a foundation for how support operates globally. Today, Fin is involved in almost every chat conversation and resolves more than 43,000 requests each month, while reshaping how support is perceived across the company.

“We didn’t just want AI that could answer FAQs,” says Yamine Gluchow, VP of Information Systems. “We needed something that could operate across products, regions, languages, and legacy tools, and deliver real business impact.”

This is a story of how an enterprise company rewired support from the inside out, not just to cut costs, but to build smarter systems, new roles, and a foundation for AI-first operations.

Lightspeed is a global commerce platform serving approximately 145,000 retail and restaurant locations. Its software powers everything from point-of-sale and payments to eCommerce and inventory, which means support has to be fast, accurate, and always on.

As the business grew, two big challenges loomed:

1. Volume: Tens of thousands of support requests each month, spanning multiple products, languages, and time zones. Most were coming in by phone or email – one-to-one interactions that were expensive to staff and impossible to scale. “When support volume gets that high, your only option is throwing bodies at the problem,” says Yamine. “It’s expensive, it doesn’t scale, and it’s not how you deliver great service.”

2. Complexity: Years of mergers and acquisitions had left a multitude of systems: numerous Salesforce instances, overlapping ERPs, and siloed data. Automating across such a fragmented environment felt daunting.

Meanwhile, expectations across the board kept rising. Customers wanted instant answers. Leaders wanted lower costs. And Yamine’s team, who own the infrastructure, data, and tooling that underpin customer-facing operations, had to deliver both, without tearing down and rebuilding the entire stack.

AI felt like a natural solution, so the team experimented with building their own AI assistant. This gave them valuable insight into just how complex and resource-intensive it would be to solve the problem on their own – so when Yamine received an email about a new beta AI Agent called Fin, he was immediately bought in: “I got the email at 2am. My response was three words: I want in,” he recalls.

That eagerness to test and learn quickly set the tone for what came next.

There’s a common perception that introducing AI into a large enterprise means months of painful integration before you see value. Fin is designed differently: it’s built to plug into existing systems, work with diverse data, and deliver results fast. Lightspeed proved just how quickly that can happen.

With support from leadership and a culture that embraces experimentation, Fin went live within weeks. “We move fast, we test, and we don’t overcomplicate,” says Yamine. “That mindset made it possible to get Fin live in weeks.”

The technical integration was the easy part. Fin fit seamlessly into a complex environment of Zendesk for ticketing, multiple Salesforce instances for CRM, overlapping ERPs, and siloed knowledge bases – all mission-critical and in constant use. Crucially, it did this without requiring Lightspeed to rip out or replace those systems. Fin worked with what was already there and proved it could deliver value inside a complex stack.

Behind the scenes, Fin further proved its ability to thrive while working as part of a complicated setup. Leveraging Salesforce entitlements, internal APIs, and custom logic, Fin segments users and handles diverse support needs across products, regions, tiers, and languages.

“It felt like cutting the Gordian knot,” reflects Yamine. “We didn’t have to untangle everything, Fin worked with our complexity, not in spite of it.”

But integration alone wasn’t enough – the real key to performance came from governance and training. “It’s not magic. If you invest in understanding, adoption, and great content, AI performance takes off,” states Yamine. That discipline was embedded from day one, as Angelo Livanos, VP of Global Support, outlines in his own story.

Within weeks, Fin was resolving thousands of requests end to end, laying the foundation for scale.

What began as a way to handle more tickets, faster, has become a core layer in Lightspeed’s customer operations.

The numbers tell part of the story:

• 88% of all support conversations now involve Fin, with 100% of chat conversations starting with the AI Agent.

• 72% of Fin conversations are resolved without human intervention.

• More than 43,000 customer requests are resolved by AI each month.

• Service is delivered in 12+ languages, across more than 100 countries.

• Crucially, CSAT has remained stable – and even improved in some markets – proving AI hasn’t come at the expense of customer experience.

What makes this remarkable is how far Fin has come since the early days. “We started at 20% resolution,” Yamine shares. “Today, we’re at over 70%. Watching that number climb was shocking. There was lots of celebration, maybe some champagne.”

And it’s not just simple queries. Fin is increasingly resolving complex, multi-step requests. Yamine recalls one striking example: a merchant in France contacted support with questions about their tax invoices. Normally this would have required a lengthy phone call with an agent, checking the back-end data and explaining the rules step by step. Instead, Fin handled the entire interaction in French, providing an accurate explanation end-to-end. “I was shook,” says Yamine. “We didn’t expect Fin to go beyond simple requests, but here it was resolving market-specific queries that not every human agent could have answered – and the customer left a positive CSAT response.” 

Fin’s success has shifted perceptions inside Lightspeed. Support has become a leader of innovation, with other GTM departments looking to it as a model for AI adoption. 

Operationally, the change has been profound. Lightspeed created a dedicated Digital Engagement team to manage Fin’s optimization, and formalized a triangular model that brings together technical teams, frontline experts, and content specialists. 

The channel transformation is equally stark. Where once the vast majority of support volume came through phone and email, today chat is the default. Fin not only accelerated that shift, it made it sustainable at global scale. “When we first rolled Fin out in 2023, there was healthy skepticism,” shares Yamine. “But, true change comes from within – people saw it working in practice, and that’s what made every skeptic a believer.”

At Lightspeed, Fin has become a foundational AI layer – handling millions of conversations, orchestrating complex workflows, and powering a new generation of customer experience.