Greetings, Fellow DaoistA greeting in Xianxia novels, between cultivators on the same path, as a gesture of mutual respect.
Keifer Lee
I build ML systems to tackle Decision Intelligence problems at scale, with a focus on high-risk domains such as fraud and abuse, security, and finance.
I'm an AI Scientist on Intuit's Trust & Safety team in New York, and a Member of Technical Research at the New York Machine Learning Research Guild (NY-MLR). Before that, I worked on robotic perception and planning at Nokia Bell Labs and at NYU's AI4CE Lab with Prof. Chen Feng, and earlier still, built credit risk systems at MoneyLion that processed and safely underwrote billions of dollars in transactions each year. I hold an M.S. in Computer Engineering from New York University, on a Merit Scholarship, and a Bachelor's in Electronics Engineering from Taylor's University in Malaysia, on the Grand Challenge Scholarship. In my free time, I like to read "proper" books and much more fun (but less healthy) light novels (write to me with your book recommendations!).
News
New technical note: A Quick Tour of GBDT & XGBoost, building both from scratch and understanding why they still work so well in practice in 2026.
New essay and report: Inference-Layer Security, on defending a model-serving platform against bot farms, model extraction, and agentic misuse. The synthetic dataset behind it, Adversarial Inference Sessions, is open on Hugging Face.
Helped design and run the Intuit TechWeek NYC AI/ML Hackathon, an SMB underwriting challenge. Thank you to everyone who took part, and to the team at Intuit NYC for making it possible!
SAVMap appears at ICRA 2026. Preprint is on arXiv.
Community
I host the New York Machine Learning Research Guild (NY-MLR), a local, informal community. Every month, our colloquium brings together researchers and practitioners at the frontier to share, disseminate, and collaborate on their work, blending academia and industry for a vibrant research scene in NYC.
Recent and upcoming sessions:
Zayne Sprague · NYU Courant, Google
From Chain of Thought to Agent Swarms: A Brief Story on Reasoning in LLMs
Arthur Jacot · NYU Courant
Deep Learning as a Convex Paradigm of Computation: Minimizing Circuit Size with ResNets
Yair Schiff · Cornell, NVIDIA
Speed, Guidance, and In-Place Editing with Diffusion Models for Text
Rohun Agrawal & Leon Li · Columbia & NYU Courant
Advances in Architectures for Long Context and Retrieval
Xinhao Liu · NYU Tandon, Amazon FAR
Grounding Over Appearance: Scaling Embodied AI with the World's Own Geometry
Interested in joining the Guild, or have an idea for the community? Reach out at keifer@nymlr.com.
Publications & Whimsical Writings
- SAVMap: Structure-Aided Visual Mapping of Large-Scale 2.5D Manhattan Wireframes from Panoramic Video
Howard Huang, Bharath Surianarayanan, Keifer Lee, Chenyu Wang, Chen Feng
- SO-NeRF: Active View Planning for NeRF using Surrogate Objectives
Keifer Lee, Shubham Gupta, Sunglyoung Kim, Bhargav Makwana, Chao Chen, Chen Feng
- A Quick Tour of XGBoost
An applied practitioner's tour of Gradient Boosted Decision Trees and XGBoost - where it came from, how it actually works under the hood, and understanding why it works so well on tabular data
- Inference-Layer Security: Defending Against Adversarial Inference and Infrastructure Abuse
An exploration on how to secure Model-as-a-Service infrastructure at the inference layer against runtime exploits, ranging from automated bot farms to adversarial extraction and agentic misuse
Reviewing
Speaking
I'm happy to give talks and guest lectures, or join panels and podcasts. My main interest is Applied Decision Intelligence: using ML to make good decisions under uncertainty and risk, through forecasting, experimentation, causal inference, and optimization. Topics I speak on:
Applied Decision Intelligence
Turning predictions into good decisions under uncertainty and risk.
Alignment, Safety & Robustness
Keeping AI systems secure and reliable, even against adversaries.
Probabilistic ML & Uncertainty
Knowing when, and how much, to trust what a model tells you.
Grounded & Verifiable Agents
Agents that use tools and model the world they act in, and evaluations that tell real progress from noise.
AI Governance & Policy
Policy for AI, and how society can coexist with, and be augmented by, increasingly capable machines.
Contact
The best way to reach me is: keifer@nymlr.com. For talks, a line on the audience, format, and date helps.