Hi, I'm Yuan 👋
Building AI/ML applications to solve real-world problems.
I work at the intersection of AI and biomedicine. Trained as an immunologist, I spent years decoding how T cells respond to viruses like dengue and SARS-CoV-2, and how they can be engineered to fight cancer. These days I build AI agents for various applications.
Seattle, WA
Featured projects
All projects →FM4Life
Structured knowledge bundles ("skills") that let AI coding agents work with life-science foundation models like ESM, AlphaFold, RFdiffusion, DiffDock, and scGPT out of the box.
- AI Agents
- Foundation Models
- Life Sciences
finetune-esm
Scalable fine-tuning of ESM-2 protein language models with distributed training (FSDP/DeepSpeed via Ray) and parameter-efficient techniques such as LoRA.
- PyTorch
- Protein LMs
- LoRA
- Distributed Training
image_tabular
Python library for training deep learning models that combine image and tabular data in a single network, built on fastai. Used for the SIIM-ISIC melanoma classification challenge.
- Deep Learning
- fastai
- Computer Vision
Experience
Anthropic
Jun 2026 – Present
Member of Technical Staff · Seattle, WA
Amazon Web Services (AWS)
Aug 2024 – Jun 2026
Applied Scientist, Generative AI · Seattle, WA
Built generative AI applications on Amazon Bedrock for enterprise customers: LLM workflow assistants, voice AI simulation with agentic tool-calling, multi-agent systems for pharmaceutical research reports, and RAG support assistants. Also developed a fine-tuning approach for protein language models for binding-affinity prediction (oral presentation, AMLC 2025).
Gilead Sciences
Jan 2022 – Jul 2024
Applied Research Scientist · Seattle, WA
Machine learning and LLMs for drug discovery research: an internal RAG chatbot serving 200+ researchers, ReceptorGPT for TCR similarity search over ESM-2 embeddings, transformer models for antibody antigen specificity, and survival-model biomarker discovery for cell-therapy response.
Fred Hutchinson Cancer Center
Nov 2019 – Dec 2021
Staff Data Scientist · Seattle, WA
Single-cell genomics meets machine learning: immune-cell correlates of reduced HIV risk after vaccination, multimodal image + tabular deep learning for melanoma detection, and a COVID-19 single-cell portal integrating 21 multi-omics datasets spanning 3M+ cells.
La Jolla Institute for Immunology
Jun 2016 – Nov 2019
Postdoctoral Fellow → Bioinformatics Scientist · La Jolla, CA
Systems immunology of human T cells in the Sette lab: discovered unique phenotypes of CD4 Temra cells, profiled the transcriptomes and TCR repertoires of dengue-specific T cells, and introduced CyTOF high-dimensional single-cell analysis to the lab. AAI Intersect Fellow bridging immunology and bioinformatics.
Education
Georgia Institute of Technology
M.S. in Computer Science
2024 – 2027 (expected)
University of Alabama at Birmingham
Ph.D. in Microbiology and Immunology
2011 – 2016
Tianjin University
B.E. in Bioengineering and Biomedical Engineering
2007 – 2011
21 research publications along the way. Browse them here.
Recent writing
All posts →Building an Agentic Resume Matcher: Python Foundations for GenAI
Learning Python foundations for GenAI development through a practical agentic application: a resume matcher.
- GenAI
- AI Agents
- Python
Building a Tool-Using LLM Agent from Scratch
Creating a UFC fight recommendation agent with tool-using and reasoning capabilities, without frameworks.
- AI Agents
- LLM
- GenAI
Building Transformer Models for Proteins From Scratch
A practical guide to building and evaluating protein language models, starting from the transformer architecture itself.
- Transformers
- Protein Language Models
- Deep Learning