Develops features on a remote, scalable, high-throughput ML inferencing service to rank relevant ads to show on Amazon’s retail website, leveraging Triton for GPU inferencing. This enables using large-scale DL models (1B+ params) for ad ranking within a 40ms latency budget at 200K+ TPS to support a $1B advertisement ecosystem.

Develops features for a microservice deployment platform used to host real-time services in Amazon Ads. This leverages ECS/EC2 + Envoy proxy to build a network of 15k+ hosts with 1ms network latency with out-of-the-box client-side load-balancing, traffic shaping, and observability (request tracing).

On the Machine Learning Delivery team within the Sponsored Products organization. Works to develop our scalable remote ML inferencing solution for the advertisement delivery pipeline. Helps maintain an inferencing server that can load and serve modeling results to our ad server.

Taught students aged 8 to 18 in the principles of computer science and programming through a series of focused curriculums and hands-on programming projects.

Manages internal relations within an organization of over 90 active members.

Coordinates regular secretariat (member) events, banquets, retreats, etc.

Creates resources for professional development and establishes a human-resources presence in the organization.