(Image source: Fisent website.)

Toronto-based Fisent Technologies applies AI to automate one of the most persistent friction points in insurance—interpreting unstructured data and information that breaks workflow continuity. “We focus on applying AI to business process automation, specifically information processing,” says founder and CEO Adrian Murray. “That’s really our whole thing. We’re narrow by design, but deeply focused on a very real problem.”

Fisent’s origin story traces back to the banking sector. Murray launched the company five years ago with a know-your-customer (KYC) automation platform that served commercial banks. But as he tells it, a revelation occurred when generative AI entered the scene in late 2022. “Everyone was thinking about generating information,” he recalls. “Our insight was that the other side of generating information is interpreting it.” The team realized that interpretation—summarizing, classifying and extracting meaning—was the biggest automation gap in KYC processes. “We solved that problem first, and soon insurers began coming to us saying, if you can do that there, you can do it for claims, submissions, endorsements and other complex workflows.”

The result is Fisent BizAI, the company’s flagship platform, developed with investment from Pega (Cambridge, Mass.), which also provided early validation. “Our original product was built on Pega, and many of our team members have Pega expertise,” Murray explains. “That made it natural for us to integrate at the point where automation breaks—where information enters the process and people still have to read, interpret and decide.

Closing the Automation Gap

Unlike platforms that promise to reinvent entire workflows, Fisent injects AI through APIs into existing systems—often underwriting workbenches, claims management tools or process automation engines—to bridge the last mile of human decision-making. “We’re not trying to replace the infrastructure insurers already have,” says COO Brent Baiotto. “We’re codifying what is today human cognition—figuring out what something is, what it means and what to do next.”

He describes the result as enabling “a contiguous journey,” eliminating the starts and stops that plague digital workflows. “When content comes in—an email with spreadsheets, images and PDFs—someone has to make sense of it before automation can resume,” Baiotto says. “We fill that gap.” The impact, he adds, is immediate: “Our customers aren’t just saving money. They’re focused on speed and quality—on time to decision and time to resolution.”

That distinction can make or break a deal. “In the Lloyd’s market, one executive told us, ‘Six minutes wins the business, 12 minutes loses it,’” Baiotto says. “That’s the competitive reality when responses depend on understanding dozens of complex documents in real time.”

From Point Solutions to Expanding Platforms

Fisent’s value lies partly in its precision and partly in its repeatability. Each engagement typically begins with a narrowly defined “micro-journey”—such as endorsements, submissions or retirement-plan enrollment forms—and expands rapidly as new opportunities emerge. “Our customers don’t want 14 point solutions,” Murray notes. “They want one solution that can move horizontally across use cases.”

He points to one Canadian insurer whose digital mailroom processes millions of forms through traditional scanning and indexing systems. “They told us there were hundreds of small problems those systems couldn’t handle,” Murray says. “We started with one—group retirement enrollment forms that varied by employer and couldn’t be processed digitally. Once we solved that, they realized we could address dozens more.”

This iterative model drives organic growth. “Every single customer we have radiates out to new use cases,” Baiotto says. “We just sent SOWs number 12 and 13 last week. Each started as one project and kept expanding.”

Fisent’s approach differs from the previous generation of document-processing solutions that required extensive model training and retraining. “BizAI can understand information it has never seen before,” Murray says. “Traditional systems needed hundreds of samples to learn where a name or value appears on a form. If the form changed, you retrained the model. But humans don’t work that way—we read and infer. That’s what our system does.”

Baiotto adds that while every vendor now claims to use AI, many still rely on static data capture. “The intelligence built into modern models is already beyond what most teams can replicate,” he says. “The challenge isn’t knowledge—it’s application. How do you harness that intelligence to achieve real, contiguous automation?”

The company’s success, Murray argues, lies in pairing this cognitive layer with practical deployment discipline. “We start every project with a proof of concept,” he explains. “Within days, clients can see their own use case fully digitized. Then the question becomes: how do we move it to production? That’s where most automation projects fail, but it’s where we excel.

Production-First Philosophy

Founded in 2021 and pivoting to its current model in late 2022, Fisent has raised $2 million in seed funding and now counts roughly 10 enterprise customers worldwide. The company reports more than 500 percent growth in usage and strong revenue gains over the past year.

Murray says what differentiates Fisent isn’t theoretical innovation but consistent delivery. “Every customer who’s licensed Fisent BizAI has gone into production or is on the way,” he says. “That’s a 100 percent success rate—compared to roughly five percent in the broader market, according to MIT.” He attributes this to a focus on outcomes rather than hype. “We’re not fantasizing about rewriting everything with AI. We’re using it tactically to close the automation gap—where previous systems stopped.”

Baiotto frames the company’s philosophy in broader terms. “Every boardroom is asking, ‘What are we doing with AI?’” he says. “But the real question is, what’s the value of AI? Is it in owning it—or in applying it to make a difference for customers, employees and shareholders? The people who focus on applying it for outcomes are the ones who will win.”

Practical AI for a Fragmented Industry

For an industry long defined by paper, email and legacy systems, Fisent’s message resonates precisely because it’s pragmatic. “Most carriers have invested heavily in modern cores and digital platforms,” Murray says. “They don’t need to rebuild. They just need help making those systems truly intelligent.”

That intelligence—applied in weeks rather than years—may be the key to bridging the gap between automation and adaptability. “We’re solving what was previously impossible or impractical,” Baiotto says. “And the market is finally ready for it.”

ITC Briefing: SAS’s Franklin Manchester on Governing AI Responsibly

ITC Briefing: Percipience’s Bruce F. Broussard, Jr. on Data for AI

Socotra CEO Dan Woods on Innovation and Insurance’s Technology Gap

ITC Briefing: EY’s Chris Raimondo on Insurers’ Enterprise Turn Toward AI

ITC Briefing: Fenris’ CEO on Predictive AI and Distribution Efficiency