Dec 2023 – Sep 2024 · 9 mos · United States

• Tech lead@Waymo Foundation Model.
• Tech lead@offboard perception foundation.
• Pioneers across products, engineering and research to help Waymo deliver the worlds best driver.
• Research projects:
• waymo E2E foundation model, tech report: https://arxiv.org/abs/2410.23262
• Product & engineering projects:
• offboard VQA system
• offboard autolabeling
• e2e perception
• vehicle semantics model/data/eval infra upgrade
• sensor fusion algorithm, onboard optimization.
• 3D detection across platforms and operation domain.

Nov 2019 – Mar 2021 · 1 yr 4 mos

• Cofounded several Vertex AI product units, including
• data labeling & dataset management: https://cloud.google.com/vertex-ai/docs/datasets/overview
• model online monitoring & eval: https://cloud.google.com/vertex-ai/docs/model-monitoring/overview
• Published two papers in ICML and ECCV

Mar 2017 – Oct 2019 · 2 yrs 7 mos

• Worked on several CloudAI early product
• AutoML vision labeling features: https://cloud.google.com/automl?hl=en
• noisy web data augmentation service

Jun 2016 – Sep 2016 · 3 mos · San Francisco Bay Area

Nov 2015 – Jan 2017 · 1 yr 2 mos · San Diego, California, United States · On-site

• One of the early interns when Tusimple was just founded. Build Tusimple’s first distributed training infrastructure and SLAM prototype.

Jul 2014 – May 2015 · 10 mos · Hong Kong

• · Designed and implemented a computation offloading system for wearable devices with Java
• · Designed and implemented a hybrid computation offloading system with Android and Java
• · Built a kernel regression based link prediction model and implemented it with Java and MySql