What I Do
My work focuses on building systems and solving problems. This can look like building pipelines on local hardware where PII is a major concern and off-site models are not an option. Other times it looks like leveraging agentic solutions to build 0-1 products for fast prototyping.
I hold an M.S. in Data Science from the University of San Francisco, where my research focused on combining spatial analysis (ArcGIS, Census data) with NLP and network graph analytics (NetworkX, LDA) to support constitutional litigation.
You can find me teaching at USF, making art in various Bay Area warehouses and surrounding woods, and in the lab researching about the internal geometries of LLMs.
Systems & Engineering Philosophy
1. Data Sovereignty & Local-First AI
PII is important, and there are countless scenarios where you cannot just throw your data into a SOTA model that is hosted offsite. I prioritize these when considering the solutions that I am designing.
2. Interpretability
Why a model (LLMs to small classifiers) performs the way it does is at worst a nice thing to know, and at best a critical aspect of the process. Regulators would argue it should always be the latter. Disruptors the former.
3. End-to-End Engineering Craft
A little bit of back end, a little bit of front end — we are in the era of a little bit of fullstack. While my skills are centered in backend concepts, my projects often require a frontend component for the UX to be a little more gentle.
Technical Toolkit
Data Systems & Pipelines
Python • pandas • DuckDB / SQL • MongoDB • Airflow • Metaflow • DVC • Tesseract OCR
Modeling, NLP & Graph Analytics
PyTorch • Transformers • Whisper • scikit-learn • NetworkX • SHAP • UMAP • ArcGIS / GeoPandas
Infrastructure & Governance
Docker • Kubernetes • GCP • MLflow • Linux / OpSec • Git & CI/CD
Community Mentoring & Visual Computing
Beyond production machine learning and research:
- Community Data Mentoring: I worked with Street Spirit in the East Bay, providing data literacy training and technical guidance for grassroots investigative reporting—including supporting the creation of Swept Off The Map to document and visualize municipal encampment sweeps.
- Generative Computing & Computational Physics: I write custom generative sketches and physics simulations in Processing (Java) and GLSL—exploring N-body gravitational dynamics, relativistic light bending, genetic cellular automata, botanical growth models, and sunset gradient fields.
- Art & Music: I enjoy the community experience of techno, so you’ll often find me helping design, build, or run various raves and parties throughout the bay.