National loyalty & engagement platform

Situation

A national loyalty and engagement platform company had the problem every successful multi-tenant product eventually earns: a monolithic platform, built years earlier, now carrying marquee enterprise clients and straining under their growth. Reporting workloads were dragging down the core database. Releases were heavyweight. And every architectural decision had to answer one hard constraint — the platform’s biggest clients could not feel any of this happening.

The internal team knew the platform needed to evolve. What they wanted was senior architectural judgment alongside them — and execution capacity that could move without pulling their own people off the product.

What we did

The engagement came to us through a managed cloud-services partner, and we ran it the way we run most of our work: Ripple owned the strategic relationship and oversight — CTO advisory working directly with the client’s SVP of technology and CTO — while our delivery teams supplied the architecture and engineering muscle. The first pass was diagnostic and structural: a domain-driven design exercise that mapped the monolith’s seams and carved out the first microservices — reporting first, since that’s where the pain was, then authentication and content management.

Performance work landed alongside the decomposition rather than waiting for it. A caching strategy combining Redis with a front-end CDN, read replicas to take reporting load off the transactional database, and a move to managed cloud queueing and storage services simplified operations while the bigger restructuring proceeded. The platform migrated to a long-term-support .NET version the same way everything else moved — incrementally, module by module, never as a bet-the-company cutover.

Phase two took the same philosophy to the riskiest part: moving clients. Infrastructure went to code (Terraform, Kubernetes), automated testing came in ahead of the migrations, and tenants moved in waves — a pilot client first to prove the path, then batches of increasing size, with the largest enterprise accounts last, after the process had been derisked on everyone else.

Outcome

The platform that strained under its tenants became one that scales with them: microservices where the monolith hurt most, reporting that no longer drags the core database, infrastructure rebuilt as code, and a client-migration machine that moved tenants in controlled waves instead of one terrifying weekend. The team we worked alongside owns it all — the architecture survived our departure, which is the test that matters.

A year after the engagement wrapped, the client’s technology leader described where the company had landed — investing in a next-generation platform and an AI-native engineering organization, built on the foundation that no longer fights back:

“We’ve been around for 25 years. With some of the steps that we’re taking with our platforms and our approaches, it’s really exciting to see how we’ve leveraged this time to position ourselves for the future.”

— SVP of Technology, client company