• ml infra, data platform: ingest, storage, orchestration of compute, etc, of PBs data / week

• Q4 2025: Designed & built an observability pipeline to capture the cost, virtual resource utilization & efficiency / parallelism of $1M/month of Apache Spark jobs on EMR Serverless / EC2.

• Q1 2026: Optimization of Apache Spark on EMR. Failure control, anomaly detection & virtual resource management.

• Q1 2026: Batch based processing for 10-20M simulated driving scenarios per week to save $70k/month in compute & decrease latency by… 🤩

• Weekly group covering prediction / planning for autonomous vehicles.

• Presented:

1) "ReSim: Reliable Simulation for Autonomous Vehicles" (Yang et al.)

2) Model-Based Policy Adaptation for Closed-Loop E2E Autonomous Driving (Lin et al.)

• My running notes here: https://tinyurl.com/4erjhn6m

• worked w/ top* ml researchers to improve their llm capabilities (* = under nda)

• Fulfillment management system team: 100 engineers, 30 micro services, orchestrate fulfillment of 4,000,000+ online orders/day to generate $100,000,000+/day

• Built a tool for 50+ devs to quickly identify slow / failing integration tests of 29 micro-services

• Developed a Spring Boot REST API to query Elasticsearch & visualized test data w/ React

• Enabled a 60 min reduction in test suite duration across 10 micro-services by exposing performance bottlenecks