I've been fascinated by the relationship between human and machine ever since I built my first robotic prosthetic limb in high school. From training neural networks that help artificial limbs walk to training artificial minds in massive datacenters, I've seen the power of AI. But working inside frontier labs taught me that we've collectively lost the plot.
We obsess over the machine and forget the human. I've always believed the machine should serve the human, the way a robotic limb lets an amputee walk. So I left the labs to pursue this mission: understanding how AI is reshaping human formation, and how personal AI can support human autonomy rather than replace it. I do this through my research at Aesona and through my writing, including a book in progress.
If you resonate with this mission, I'd love to chat.
My Work
Aesona — Founding Researcher
- Building personal AI as a trusted advisor — in service of you, not a replacement for you.
- Entirely owned by you only.
South Park Commons — Member, Independent Researcher
- Studied how AI shapes our inner lives and what an ideal human–AI partnership looks like, hosting writing and discussion circles
Anthropic — Member of Technical Staff — Post-Training & Applied AI
- Led custom fine-tuning for enterprise and agentic-startup customers
- Co-built evals, agentic harnesses, and RL environments for the production model — working across the stack from research to commercialization
- Ran onboarding sessions on how models are trained end to end
Meta / PyTorch — AI/ML Engineer
- Co-founded torchtune, PyTorch's open-source library for post-training LLMs
- Co-founded torchmultimodal, PyTorch's library for training vision-language models
- Scaled post-training infrastructure for research and product teams across Llama 2 through 4
Perceptronics Solutions — AI Scientist
- Pioneered adversarial ML models for electronic warfare, achieving 80% success against radar jammers
- Built AI/ML decision-support systems for high-stakes defense and security environments
Stanford University — NSF Graduate Research Fellow, Computational Psychiatry
- Modeled brain networks underlying autism using graph theory, linear dynamical systems, and AI
- Collaborated with SRI International using generative vision models to reconstruct (denoise) neuroimaging data
Shirley Ryan AbilityLab — Machine Learning Researcher
- Architected deep learning models using EMG and mechanical sensor data to predict walking modes for robotic limbs
Contact
Always happy to advise on a project or chat about AI and the human condition. Reach out.