Yang.

Building agricultural intelligence through human-guided robot learning: turning demonstrations and interventions into robot task policies. Leading technical direction and prototype development.

Designed enterprise GenAI architectures for European customers using Qwen LLM serving, vision-language models, RAG knowledge bases, vector search and multimodal data pipelines. Alibaba Cloud Certified Solutions Architect Expert.

Owned the sensor-data and closed-loop resimulation pipeline for autonomous driving. Built ROS data collection, logging and validation workflows, natural-language edge-case search, and a Kubernetes GPU evaluation platform. The in-house platform reduced AWS GPU costs by 60%.

Shipped a production C++ radar and camera perception module for Motional’s Level-4 robotaxi. Built an ML evaluation pipeline and Automotive SPICE testing framework, delivered motion-prediction work, and guided sensor-fusion tracker development.

Built backend systems for an ML recommendation engine serving over six million active users in Dublin and Seattle. Led migration of recommendation pipelines from Hive to Spark, reducing runtime and storage footprint.

Built a ROS-based autonomy stack for an autonomous campus shuttle, including path planning and GPS/LiDAR fusion. Established ROS bag logging and replay for sensor data, and contributed to the live YC Demo Day demonstration.

Dissertation in robotics: multi-robot motion planning and coordination algorithms.

Electrical engineering studies.

Electrical engineering studies.