Banyan

Banyan is a receipt data platform that connects item-level purchase data between retailers and financial institutions. I joined in the early days, before the $43M Series A, as a fractional CTO and senior AWS cloud architect, and stayed on for a bit over two years.

The core of the work was an event-driven infrastructure to ingest and process purchase receipts at scale: message queues feeding Lambda functions, object storage for the raw data, relational databases for the structured side, and a search layer on top, handling billions of itemized transactions. Third-party retailer data came in over an SFTP server with an automated pipeline that normalized and stored everything without a human in the loop.

Security and compliance were a first-class requirement, not an afterthought. I led the SOC 2 Type 2 compliance initiative, architecting the infrastructure controls and the audit documentation behind it, and hardened the network by moving the databases from public subnets to private ones behind a bastion host, eliminating direct internet exposure.

I also built and managed the internal developer documentation site, so the growing engineering team had one place to learn the integration patterns.