Instead of a risky 'boil the ocean' replacement, Unum modernizes legacy mainframes by first separating the customer experience layer. They use APIs to access core data, allowing front-end innovation while proving out a de-risked strategy for broader transformation.
To avoid low-value tech projects, Unum maps all investments to specific, business-owned "value streams." This structure forces the company to prioritize work based on expected business outcomes and to say "no" to lesser initiatives, shifting from an IT-centric to a value-creation mindset.
To make "AI Ready" tangible, Unum uses a two-pronged approach. "Everyday AI" (e.g., Copilot) is rolled out to the entire company to foster citizen development and reduce fear. "Embedded AI" involves deep, mandatory training for engineers to integrate AI directly into their core workflows and boost productivity.
In its 175-year-old culture, Unum found a passive 'pull' strategy for AI training (making it available) failed due to low adoption. They shifted to an active 'push' strategy with incentives and an 'AI champion' program to drive engagement, overcoming the inertia where a 'build it and they will come' mindset falls short.
To modernize its claims process, Unum first used AI to analyze 1.5 million lines of legacy COBOL code. They then fed this analysis into Pega Blueprint, effectively reverse-engineering the embedded business logic to visualize and create reimagined, modern workflows as a starting point for their transformation.