Designed, built and managed a first-of-its-kind payment rail within our personal finance application - moved >$30 million for our users to pay recurring bills and build credit

Developed a card network authorizer backed by a JIT card/ACH/ledger payment rail with high uptime and latency requirements to ensure customer bills were paid on time

Maintained a ledger of accounts for 10s of thousands of users as money flowed from bank accounts to Cushion to merchants - passed bank audits

Integrated with multiple user identification/security vendors to produce a verifiable KYC record for our card-holding users - passed all security audits run by sponsor banks, vendors

Managed many stakeholders: vendor partners (Lithic, Stripe), engineers (ML/AI team), Financial Institutions, to craft a system that would run short-term credit risk operation for JIT ACH funding, pass bank ledger audits and KYC security audits

Built an authenticated RESTful API system, which we treated like a B2B partner to our front-end systems to provide financial services, data to thousands of users a day

Produced integrations with Gmail and Yahoo mail OAuth systems, as well as IMAP systems - this provided data for in-house LLM optimization and training and unique user data

Ensured near 100% uptime and reliability through robust monitoring and infrastructure-as-code best practices (Terraform)

Contributed across all of our systems: Model/LLM traffic serving and managed the performance and code of a team of 4 contractors on top of our in-house engineering staff

Developed in-house systems that would take username and passwords from users and scrape into US financial institution websites to interact with customer support portals

Hired many engineers, interviewed hundreds and mentored junior engineers

Streamlined data ingestion pipelines with Kafka and Google PubSub, ensuring data processing throughput for over 100 million users

Integrated robust encryption, HSM solutions, and external APIs behind gateway layers, bolstering privacy compliance and maintaining operational consistency

Delivered scalable, high-performance data platforms to handle high-volume requests, achieving near-zero downtime and safeguarding sensitive user information

Engineered Kubernetes-based micro-services in Scala using functional programming and category theory, to reduce code complexity and improve testability