interested in interesting things
I'm a Privacy Engineer at Google. I'm interested in AI security and privacy, and want to keep learning how to make AI systems safer, more private, and more trustworthy.
New
Keystroke-DP
A live keystroke dynamics demo that turns a typing session into a profile and shows differentially private aggregate analytics. Try it here.
Currently trying to answer these questions
- Does the model really know it is a model?
- What is the understanding of the model about itself?
- Are the model's answers about “escaping” real or inspired from human text?
- What are better questions to think about?
Experience
- Google: helping build privacy-preserving Ads systems
- Enkrypt AI: built LLM guardrails and internal benchmarks
- Skyflow: 2× intern, secure enclave and transient data
- IBM: intern, secrets vault development
Selected papers
- Efficacy of the SAGE‑RT Dataset for Model Safety Alignment: A Comparative Study NeurIPS Workshop · 2024
- Robust recovery of adversarial examples ICML AML Workshop · 2024
- Privacy‑preserving keystroke analysis using fully homomorphic encryption & differential privacy ICML TPDP Workshop · 2021
Coffee Corner
Setup
AeroPress + Prismo + Fellow Opus.
Favorites
- Ritual Coffee — Sweet Tooth
- George Howell — Dota
- Saint Frank's — Little Brother
- Blue Tokai — Vienna
For fun
- I enjoy playing chess (≈1900 bullet, ≈1700 blitz and have fun playing the Scandinavian Defense sometimes :))
- I'm also interested in philosophy, especially Stoicism.
- Building blindfoldchess.ai to train.