PIER SolutionsA client's legacy JNI-based payment processing solution had become opaque, expensive to maintain, and difficult to adjust. Rather than a risky big-bang rewrite, they engaged us to evaluate, prototype, and scope a multi-stage modernization: wrap the existing JNI logic, migrate it to the web, and eventually phase it out entirely.
I co-led the proof-of-concept phase, building a lightweight service layer that exposed the JNI payment functions over a REST API. I also added interfaces to the Android client that could redirect calls through the remote JNI service when a feature flag was enabled, allowing for controlled, incremental testing without disrupting the existing production flow. Throughout the process, I validated that critical performance and compliance requirements — transaction latency, security, and accuracy — held up under the new architecture.
The results confirmed the approach was viable: remote calls added only ~15ms of latency, well within the client's acceptable threshold, and the solution passed all security and accuracy checks. I presented these findings directly to the client's senior leadership, who used the vetted proposal to pursue budget approval for the full migration the following year. While leadership ultimately did not allocate funding for the next phase, our team met or exceeded every goal defined for the engagement's scope, delivering a de-risked, data-backed roadmap for modernizing a critical piece of payment infrastructure for a major POS system.
