At Scalefocus, I own the merchant reconciliation service, defining its TypeScript and Node.js boundaries and PostgreSQL data model. I moved asynchronous settlement processing to Apache Kafka, handling roughly 2 million events daily.
I designed a reusable Jest harness to replay production-like payment events and catch settlement edge cases before release. I also added Redis caching that reduced typical payment-method lookup time from 240 ms to 130 ms.
At Tide, I built a business-account transaction API and refactored payment-state handling into Express modules, reducing repeat support escalations by 18%. I also created a Python reconciliation job and used AWS Lambda and Amazon S3 to replace a manual partner-file workflow.
Earlier at SoftServe, I developed Python ingestion jobs and Apache Spark transformations for retail analytics, processing roughly 80 million event rows per day. At Paysafe, I built JavaScript features for online payment flows and REST integrations for merchant account updates.

