How Libra Internet Bank Cut Test Creation Time by 35% with UiPath

Libra Internet Bank operates at the intersection of financial services and technology, serving niche markets including professionals, real estate investors, and agribusiness clients through a digital-first banking model. As delivery demands scaled, the bank’s software testing practices became a constraint — one that required a different approach to remain competitive and maintain the release quality its customers expect.

Before adopting UiPath Test Cloud, Libra relied on manual test case creation in Xray and Jira, with smoke and regression testing also conducted largely by hand. Some teams used Selenium and Appium for automation, but coverage was uneven and had not scaled across the organization. The gap between manual effort and the pace of software delivery was widening.

Libra piloted UiPath Test Cloud with two client onboarding teams, integrating AI-powered test case generation directly from Jira business requirements. The implementation connected Jira, Xray, Planview, and UiPath Test Manager into a unified workflow, and incorporated the bank’s existing UiPath automations into a test-case-oriented model. AI-generated test cases were evaluated against manually written ones in Xray, with strong results. Following the pilot, Libra extended AI test case generation across all project teams and launched a parallel initiative to expand smoke test automation to teams that had previously relied entirely on manual testing.

The measurable outcomes were substantial: test case creation time fell by 35% bank-wide, over 60% of all test cases are now AI-generated, and more than 90% of smoke tests in UiPath-enabled teams are fully automated. Release confidence improved across the board as smoke test suites shifted from manual to automated. The change also had a cultural dimension — business analysts, developers, and business teams gained direct awareness of how AI can support software delivery, and the consistency of Jira requirements improved because they now feed AI-generated test assets.

Libra plans to continue expanding the initiative into performance testing, self-healing automated tests, and agentic AI applications in other areas of software development and business operations. What began as a focused pilot to evaluate AI test case generation has become the foundation of a broader software testing transformation across the bank.