Ashan Panduwawala

Blueprints team

Developed the Scale RL Environments website, a platform featuring high-fidelity simulated worlds designed to train and evaluate AI agents on complex, real-world workflows.

Architected and built the frontend for a Personal Information Management platform used by 3M+ employees.

Part of the launch team for a new way of selling games on Amazon.com. The stack included a React / TypeScript frontend, and a GraphQL / Kotlin backend. Built the checkout system (frontend + backend) which handles purchasing and retrieving customer billing information.

Implemented a customer facing feature which allowed an Envoy proxy to intercept and manage DNS queries made by an application.

AWS App Mesh

Under Professor Mohamed Teli, collaborated with students Hyekang and Krishan to build a real-time driver drowsiness detection system using convolutional neural networks. Trained models to detect blinks, yawns, and driver actions, assessing drowsiness by analyzing yawn frequency and blink duration.

• Coursework involved Python, statistics, machine learning, and data visualization • Hosted 4-5 office hours/week

Transformed high-volume stock market event data received by exchanges such as the NYSE and NASDAQ into information to be used by internal consumers. Utilized Scala, Spark, and SQL to perform ETL (extract, transform, load) on the event data.