EXPERIENCE

Work on the left, education on the right, one shared spine. Condensed highlights — the full record is on the CV.

1. FEB 2026 — PRESENT

   Ambient.AI

   Software Engineer, ML Infrastructure

   • Migrated an 11-model real-time CV pipeline from monolithic PyTorch to NVIDIA Triton, enabling independent GPU memory management and concurrent execution.
   • Cut multi-model inference latency 33% with a pipeline-parallel scheduler, and tensor transport 18× via shared-memory transport.
   • Reduced cloud inference cost 80% with scale-to-zero autoscaling for idle model replicas.
2. JUN 2025 — SEP 2025

   GenseeAI Inc.

   Research Intern, AI Systems & Infrastructure

   • Architected a sandboxed AI agent execution environment on Docker and AWS ECS/ECR, cutting container startup latency 40%.
   • Co-developed a full-stack agent serving platform with multi-turn sessions, tool execution, and per-session usage tracking.
3. SEP 2024 — DEC 2025

   UC San Diego

   M.S. Computer Science

   • Graduate researcher at Shang Data Lab and the Swartz Center for Computational Neuroscience.
4. JUN 2024 — AUG 2024

   Taiwan AI Labs

   Software Engineer Intern, Algorithm Team

   • Improved Truku translation quality 14% (BLEU 28 → 32) and cut training time 13% with Dynamic Data Selection.
   • Reached BLEU 37.5 on medical-domain translation via retrieval-augmented data augmentation.
5. SEP 2019 — JUN 2023

   National Tsing Hua University

   B.S. Computer Science · Hsinchu, Taiwan