Amazon.com Inc.
Applied Scientist II
Nov. 2023 – Present
- Develop task-oriented multi-turn dialogue systems using LLM in Amazon Customer Service Conversational AI team.
- Work on Responsible and Trustworthy AI.
- Develop content quality moderation models for LLM outputs.
TikTok Inc.
Machine Learning Engineer
June 2023 – Nov. 2023
- Build machine learning models to predict user behaviors in the TikTok app search session.
- Launched the inner flow search pipeline to greatly improve the TikTok search engine user experience, leading to a significant increase of full-App-end User Stay/Duration Time.
Amazon.com Inc.
Applied Scientist Intern
May 2022 – Aug. 2022
- Developed a novel non-autoregressive phoneme-augmented Transformer model to correct language error in ASR transcriptions, which results in more robust inputs for downstream conversational bot models.
- Accelerated the inferencing time up to 6x times compared to Seq2Seq autoregressive model, improved accuracy compared to SOTA non-autoregressive ASR error correction models.
- Published paper here.
IBM T.J. Watson Research Center
Research Intern
May 2021 – Aug. 2021
- Deployed an intelligent simulation workflow on the IBM Cloud HPC Cluster with Docker and Enroot, provided the best practices to optimize the performance for containerized applications.
- Published blog could be found here.