Chime · Software Engineer
Jan 2026 — Present · San Francisco, CA
Leading a new backend platform for large partner transactions, and the credit furnishing platform behind every Chime liquidity product.
- Led development of a 0→1 backend platform supporting $100M+ in high-value partner transactions, owning core APIs and business logic for eligibility, transfer, and pledging workflows.
- Designed and launched a new credit furnishing platform for all Chime liquidity products, turning a manual, multi-day reporting process into an automated pipeline serving millions of members.
- Partnered with Legal, Compliance, and 3 engineering teams to re-architect backend services and reporting logic for Chime’s new line of credit product, unblocking its launch ahead of schedule.
- Strengthened production readiness by building an AI-powered on-call system that triages alerts and surfaces root causes and fixes in real time, saving the team 10+ engineering hours per week.
- Problem
- Chime needed a backend for large partner transactions, and there was nothing to build on yet. Credit reporting for its liquidity products ran on a legacy process spread across a lot of files. It was complicated, slow to debug and slow to add new features to. A new line of credit also couldn’t launch until its backend and reporting were reworked.
- What I built
- I led the new platform from the start and own its core APIs. They check who is eligible, handle transfers and manage pledges. I designed and shipped a new credit furnishing platform to replace the legacy one, and it now reports for every Chime liquidity product. For the line of credit, I worked with Legal, Compliance and three engineering teams to rebuild the backend services and the reporting logic. I also built an AI assistant for whoever is on call. It triages alerts as they come in and points to the likely root cause and fix.
- Scale
- Over $100M in partner transactions, and credit reports for millions of members.
- Result
- Credit reporting now runs on one platform that is easier to debug and to build on. The line of credit launched ahead of schedule, and the on call assistant saves the team more than 10 engineering hours a week.
Ruby on Rails · PostgreSQL · Amazon SQS · Airflow
Capital One · Software Engineer
Feb 2022 — Sep 2025 · San Francisco, CA
Real time decisioning and fraud detection services, and a subscription detection pipeline built with data science and product.
- Developed real-time decisioning services (Java, Spring Boot, gRPC) to process millions of daily transactions with ML models, driving a 10% increase in user spending.
- Led cross-functional initiatives with data science and product teams to build a subscription detection pipeline, identifying $10M+ in overlooked revenue annually.
- Leveraged AWS (EC2, ECS, S3, Fargate) to deploy and scale fraud detection services, implementing CI/CD pipelines to cut build times by 40% and enable faster, more frequent releases.
- Resolved critical production latency issues by implementing Redis caching, improving response times by 35% and eliminating timeouts during peak hours.
- Problem
- Millions of transactions came in every day, and each one needed a decision from our ML models in real time. The fraud detection services had to keep up with that volume. At peak hours, slow responses were causing timeouts in production.
- What I built
- I built the real time decisioning services in Java with Spring Boot and gRPC. They send every transaction through our ML models. I led a subscription detection pipeline together with data science and product. I deployed and scaled the fraud detection services on AWS (EC2, ECS, S3 and Fargate) and set up CI/CD pipelines for them. When latency got bad in production, I added Redis caching to fix it.
- Scale
- Millions of transactions a day, each one decided in real time.
- Result
- User spending went up 10%. The subscription pipeline found over $10M a year in revenue we were missing. Builds got 40% faster, responses got 35% faster, and the timeouts at peak hours went away.
Java · Spring Boot · gRPC · AWS · Redis
Tech4Good · Software Engineer
Sep 2019 — Dec 2020 · Santa Cruz, CA
Causeway, a platform that teaches web development. I worked on its design and architecture, its coding exercises and an accessible frontend.
- Directed design and architecture sessions for Causeway, a web development learning platform, leading to improvements that secured over $200,000 from competitive STEM education grants.
- Created data structures and algorithms coding exercises using Java and REST APIs for Causeway’s backend infrastructure, improving student engagement and course completion rates by 25%.
- Collaborated with UI/UX designers to design and operationalize a user research workflow, enabling structured feedback collection to inform future feature development.
- Implemented responsive frontend components using Angular and TypeScript to improve usability and accessibility of the Tech4Good platform.
- Problem
- Causeway is a platform that teaches students web development. It needed a clear design and architecture, coding exercises students would actually stick with, and a frontend that was easy to use and accessible.
- What I built
- I ran the design and architecture sessions for Causeway. I wrote data structures and algorithms exercises backed by Java and REST APIs. I worked with the UI/UX designers to set up a user research process, so student feedback could guide what we built next. I also built responsive frontend components in Angular and TypeScript.
- Scale
- A learning platform funded by competitive STEM education grants.
- Result
- The work helped win over $200,000 in STEM education grants, and student engagement and course completion went up 25%.
Java · REST · Angular · TypeScript
Bachelor of Science in Computer Science, University of California, Santa Cruz (Sep 2018 — Dec 2021) · AWS Certified Solutions Architect – Associate