For a platform designed to optimize bank pricing in real time, it was an uncomfortable contradiction: even the smallest change to a lending rate or pricing condition could take up to six weeks to go live.
This North America-based fintech, serving top-tier banks with AI-powered pricing tools, found itself caught between market expectations and legacy architecture. While customers expected instant responsiveness, the provider’s own systems remained bound to manual release cycles, code-heavy processes, and sprawling technical debt.
Something had to give.
A Platform Held Hostage by Its Own Complexity
The fintech’s platform was respected in the market. It offered differentiated products that enabled intelligent, dynamic pricing for mortgages, deposits, and personal lending.
But inside, the logic that powered these products was rigid. Business rules, rate conditions, and exception criteria were all hardcoded. Any update—however routine—required developer intervention, full test cycles, and code deployments.
The backlog grew. Engineering teams struggled to manage over 1,000 unresolved technical issues and more than 100 outdated components scattered across five siloed workspaces. And while each bank operated on its own instance, none could make configuration changes independently.
Roadmaps stalled. Client frustration grew. Four years passed with no breakthrough.
Rewriting the Rules—Without Touching the Code
Persistent was brought in to address what the client couldn’t fix internally. Using SASVA™, its Generative AI-powered engineering platform, the Persistent team restructured how configuration worked—starting with a simple premise: business users should never need to rely on code to control pricing.
The transformation unfolded in two focused phases:
Phase 1: Dynamic Configuration Framework
- Business logic was decoupled from code
- A real-time admin UI was introduced for pricing updates
- Configuration changes no longer triggered deployments
Phase 2: Platform Unification and Tech Debt Remediation
- 1,000+ technical issues were resolved
- 100+ components were modernized
- Five fragmented workspaces were consolidated into a scalable architecture
What once took six to eight weeks could now be completed in minutes—without writing a single line of code.
From Developer Dependency to Business Control
The result wasn’t just a faster release process. It was a complete change in how the platform was used.
Banks could now independently manage their pricing conditions—adjusting rules, testing logic, and rolling out updates without provider-side engineering support. Internally, onboarding new developers dropped from nearly two months to under two weeks. And with SASVA automating release planning and test cycles, engineering capacity was redeployed toward product innovation.
Previously multi-week cycles were now completed in minutes, directly from the admin panel.
The numbers made the impact clear:
- Release cycle time reduced by 50%
- Delivery costs cut by 45% versus internal estimates
- 40+ developers scaled using SASVA
- Business teams gained full pricing autonomy
According to McKinsey, enterprises that embed Generative AI in software development pipelines see a 40% reduction in time-to-market and a 50% improvement in engineering throughput. This client now operates at those benchmarks—by design, not by exception.
Unlocking What Had Been Stalled
With the core issue resolved, the fintech could finally evolve. What was once a reactive platform is now proactively extensible. The same configuration engine now enables:
- Smart autocomplete to guide rule creation
- Embedded analytics for real-time performance insights
- ML-based pricing recommendations
- Modular reuse across banking clients
The platform was no longer a bottleneck. It became a framework for intelligent change—built to evolve with market needs and user behavior. As Gartner projects, by 2026, 30% of enterprise platforms will automate more than half of their network activities, with configuration-first design playing a central role. This fintech is already on that curve.
Why It Worked: Execution, Not Just Engineering
Persistent didn’t just deliver tooling. It brought focused execution aligned with product leadership. The transformation was co-developed, not handed off. Milestones were met at 40% faster velocity than internal forecasts. Most importantly, the engagement returned control to the users who shape pricing strategy—not just the engineers who maintain it.
What had frustrated teams for four years was reimagined and resolved in six months.
Want to Build a Platform That Evolves at the Speed of Your Business? Talk to Persistent’s SASVA Engineering Team
