who i am
I’m an engineer and a tech lead, and I enjoy building creative products and solutions
01
a little about me
what drives me
I love turning ambiguous, high-stakes, consumer-facing business problems into clean, reliable technical systems
what i’m good at
Not to toot my own horn, but I’m great at driving cross-functional alignment and getting the low-level technical details right without losing sight of the bigger picture
02
Experience
- serve as tech lead for the gift cards platform, a $1.08B annual business. i align the roadmap with partners, set technical direction, and help engineers architect the systems behind team projects
- currently architecting a third-party gift card redemption product: a first-of-its-kind capability for ecommerce platforms that lets customers redeem third-party value within the commerce experience, with $225M+ in annualized revenue potential
- introduced the platform infrastructure that integrates virtual products into DoorDash’s existing physical-product retail stack, enabling the launch and scale of a gift card marketplace that brings in roughly $600M annually
- designed and built the foundations for a new pricing type that enables consumers to set the price for gift cards during checkout. this is now being reused by other teams for similar variable-price offerings, such as donations
- lead fraud initiatives for a rapidly growing, high-risk product by collaborating with the fraud team to build systems that actively monitor and identify malicious behavior, and block fraud actors without significantly hindering rapid customer growth
- built AI-led internal tools and processes that improve team efficiency, including a customer support responder and automated reliability reporting for organization-wide engineering health
- Designed and implemented core backend infrastructure for Twilio’s RBAC platform, resulting in two pending patents for identity and authorization innovations
- Led engineering efforts to migrate production databases from EC2-hosted infrastructure to managed storage with zero downtime and zero customer data loss
- Established operational ownership for six mission-critical services with 99.999%+ SLOs, improving production reliability, observability, and deployment readiness
- Contributed to regionalization of Twilio Identity services to support global regulatory and geographic requirements
03
Academic projects
GitHub ↗Scene recognition
Built a generalized deep-learning model for recognizing background scenes in images, reaching 90% accuracy on the 15-scene dataset with minimal training.
Behavioral analysis
Improved model performance by 6.7% for detecting behavioral patterns from video-recording visual cues, in support of autism-onset prediction research.
Image captioning
Trained an LSTM image-captioning model using image encodings from pre-trained CNNs, achieving a BLEU score of 0.31.
Human activity recognition
Built a model that recognizes human activity from raw accelerometer, IMU, and full-body sensor signals, with a 0.73 F-score.
Script generation
Designed a word-embedding language model that generated dialogue through beam-search decoding.
Stock market prediction
Developed a Q-learning model that used technical indicators to determine buy and sell actions.