Introduction
The customer is a leading Canadian financial services company, an insurer serving millions of policyholders across life, health, and wealth products. Its digital properties, quoting tools, policy servicing, claims portals, are the front door for customers and advisors alike, so uptime and correctness aren’t optional: a broken quoting flow or a claims form that silently fails is a customer the company can’t serve.
Like most enterprises at this scale, the QA team had far more critical journeys to protect than the hours in a week to test them by hand, and the gap between what shipped and what got properly checked was widening.
The Challenge
Production is the least forgiving place for a regression to surface. Quoting, policy management, and claims each span multiple products and customer segments, and every release could touch any of them. Manually walking through that surface before and after every deployment simply didn’t scale, and by the time a manual tester caught an issue, it had often already been live for hours.
The team needed continuous, production-grade test coverage, but without months of setup, custom scripting, or a large dedicated automation team to build and maintain it.
The Solution
From kickoff to coverage in a single day
The company adopted Klarent’s autonomous QA agents and had them writing and running tests against production almost immediately. Rather than a long onboarding project, the team described its critical journeys in plain language and had them running as scheduled, autonomous test suites before the week was out.
150+ production tests automated on day one.
In a single day, the team stood up more than 150 tests covering its production environment, work that would typically take a traditional automation effort weeks to scaffold.
Continuous, scheduled coverage.
Once live, the suite didn’t sit idle waiting for the next release. Klarent runs 100+ automated test executions every hour, so the company’s most important flows are being verified around the clock rather than at the cadence of a release calendar.
No specialist bottleneck.
Because each test is written in natural language, the team could describe new scenarios and get them scheduled without waiting on a dedicated automation engineer, which is what made same-day coverage possible in the first place.
The Impact
Value realised in the first week
Speed of adoption is only useful if it translates into real coverage, and here it did immediately: within 7 days of going live, the always-on test suite had already caught genuine bugs in production, issues that, without continuous automated checks, would likely have been found by a customer instead of a test.
That first week set the pattern for how the company now thinks about QA: not a gate that runs occasionally before a release, but a constant presence watching production, catching regressions close to the moment they’re introduced rather than after they’ve had time to cause damage.
What’s Next
Building on the fast start, the company is expanding its Klarent test suites deeper into policy servicing and claims, growing the 150+ initial tests into broader coverage across its digital estate. The bar the team has set for itself is simple: keep the time from “we described this journey” to “it’s under continuous, automated watch” as close to zero as it was on day one.
Note: Customer details have been anonymised at the customer’s request.
