CV
A copy of my CV can be downloaded here.
This file was last updated on Thu Mar 19 2026.
Bio
NEW
Learn More →
I wrote a book on AI agents!
I am a Research & Engineering Lead at Microsoft Core AI, where I build AI agent infrastructure and developer tools. I am a core developer for AutoGen (52K+ stars), Microsoft Agent Framework (5.9K stars), and a contributor to Azure Foundry. My research spans AI agents, developer tools, and human-AI interaction, published at ACL, EMNLP, AAAI, and CHI with multiple best paper awards. I am an IEEE Senior member and Google Developer Expert in ML.
Experience
Microsoft Core AI
May 2025 - PresentResearch & Engineering Lead · Santa Clara, CA
- Created Agent Optimization Service from scratch — now serves hundreds of thousands of customers on Azure.
- Led unification of Semantic Kernel and AutoGen into Microsoft Agent Framework. Created DevUI (470K+ downloads), featured in 6+ Ignite 2025 sessions.
- Led UI direction across 3 product surfaces (Agent Framework OSS, VS Code Extension, Azure Foundry). Co-authored A2A on MCP documentation (Top 5 DevBlog).
Microsoft Research - AI Frontiers Lab
Oct 2021 - May 2025Principal Research Software Engineer · Sunnyvale, CA
- Core developer for AutoGen (52K+ stars, 5.9M+ downloads). Co-authored Magentic-One, industry reference architecture for autonomous multi-agent systems.
- Created LIDA (3.2K stars), an automated visualization framework using LLMs. Approaches adopted by Excel, Fabric, PowerBI teams and Project Sophia.
- Improved offline evaluation metrics for GitHub Copilot (14% higher correlation with customer satisfaction) used to select models for millions of users.
Cloudera Fast Forward Labs
Jan 2019 - Sept 2021Principal Research Engineer · New York / Palo Alto
- Led research reports on Deep Learning for Question Answering, Anomaly Detection, and Image Analysis. Built NeuralQA, an extractive QA library using BERT.
- Led development of Applied ML Prototypes (AMPs) for Cloudera ML — became the standard tool for customer onboarding.
IBM Research
Apr 2016 - Jan 2019Research Staff Member · Yorktown Heights, NY
- Created Data2Vis, the first and most cited neural network approach to automatic data visualization (IEEE CG&A Best Paper, IEEE VIS Best Paper Honorable Mention).
- Co-created TJBot, an open-source DIY AI kit adopted by 8,000+ users for classroom teaching and corporate training.
Education
City University of Hong Kong
2012 - 2016PhD in Information Systems (Quantitative User Behaviour, HCI)
Carnegie Mellon University
2009 - 2011MSc Information Networking
Skills & Open Source
Languages & Frameworks
Python, TypeScript/JavaScript, React, Node.js, PyTorch, LLMs, Multi-Agent Systems
Open Source Projects
- AutoGen — 52K stars
- Agent Framework — 5.9K stars
- LIDA — 3.2K stars
- Handtrack.js — 2.9K stars
Awards & Patents
Awards
- Designing Multi-Agent Systems — Amazon Best Seller, #1 New Release in AI (2024)
- Best Paper Award, IEEE Computer Graphics & Applications (2020)
- Best Paper Honorable Mention, IEEE VIS (2018)
- Best Technical Demo, AAAI (2018)
- Heidelberg Laureate Forum — 1 of 200 young researchers invited (2018)
- Grand Prize, #BuiltWithTensorflow Challenge (2019)
- IBM Open Source Award (2017)
Patents
7 granted, 2 pending.
- Automated Generation of Data Visualizations Using LLMs and Diffusion Models — US 12,518,447 (Granted 2026)
- Data Health Evaluation Using Generative Language Models — US 12,579,115 (Granted 2026)
- Embedded Attributes for Modifying Behaviors of Generative AI Systems — US 12,423,338 (Granted 2025)
- Coding Activity Task (CAT) Evaluation for Source Code Generators — US 12,254,293 (Granted 2025)
- Detecting Human Input Activity Using Wearable Inertia and Audio Sensors — US 11,195,118 (Granted 2021)
- Automated Summarization Based on Physiological Data — US 10,353,996 (Granted 2019)
- Using Ultraviolet Sensor Data to Determine a Pseudo Location of a User — US 10,066,987 (Granted 2018)
- Pending:
- Using Large Generative Models to Improve the Performance of Weak Language Models — US App. 2025/0348745A1
- Applied Machine Learning Prototypes for Hybrid Cloud Data Platform — US App. 2023/0267377A1
Talks
A list of talks I have delivered.
- October 2024
CfD Conversations SeriesHybrid (In-person & Virtual).
What Can GenAI Really Do for Data Visualization? - December 2023
Ai.dev / Cassandra SummitSan Jose, McEnery Convention Center.
AutoGen: NextGen AI Applications via Multi Agent Conversations - June 2021
Fast Forward LiveOnline.
Deep Learning for Automatic Offline Signature Verification - November 2019
QCON SFSan Francisco, USA.
ML in the Browser: Interactive Experiences with Tensorflow.js - November 2019
Tensorflow World Santa Clara, California USA.
Handtrack.js: Building gesture-based interactions in the browser using TensorFlow.js - November 2019
Google ML SummitPittsburgh PA, Cambridge, USA.
Art + AI : Generating Novel African Mask Art using Generative Adversarial Networks - September 2019
OReilly Strata Data Conference New York, USA.
Handtrack.js: Building gesture-based interactions in the browser using TensorFlow - May 2019
!!Con New York, USA.
Dance of the Ancestors: I used Neural Networks to Re-imagine African Mask Art !!
Read the Newsletter.
Read and Subscribe
Powered by Substack. Privacy Policy.
I write a monthly newsletter on Applied AI and HCI. Subscribe to get notified on new posts.
.