Data Scientist and AI systems builder based in Waukesha, WI. I recently completed the TripleTen Data Science program, and alongside it I have been building a full AI automation stack for a real e-commerce business.

I come from an unconventional background: food service, retail, social work, and a B.S. in Criminology with a Psychology minor. That mix taught me to read people, understand how systems behave, and ask the right question before I touch the data.

Python, SQL, pandas, NumPy, scikit-learn, PyTorch, Plotly, Streamlit, Modal (serverless GPU), and LoRA / DPO fine-tuning.

Real systems I design and build, not course exercises.

Continuum Goods AI Stack (private, available on request) An autonomous multi-agent system that runs a live Shopify store: pricing, inventory, fulfillment, trend research, listing generation, marketing, and email, all coordinated by a master scheduler. It uses a self-improving "agent brain" design where behavior is tuned by editing plain text files instead of code. A sanitized public slice lives here: continuum-agents.

Orlog (part of the stack, private) A multi-model AI decision engine. It routes a question through several independently reasoning models and synthesizes a well-rounded answer, backed by a custom DPO training pipeline for distilling that reasoning into a single fine-tuned model. This is the work behind my writing on why an AI system needs an outside observer to stay honest.

PersonaVid An end-to-end AI video personalization pipeline: footage cataloging, shot direction, GPU video generation on Modal, compositing, quality scoring, and publishing, run as one connected system.

Algorithmic Trading System (private, available on request) A multi-timeframe trading system with minute, hour, and week agents, a decision council, risk management, and a trade executor.

TripleTen Data Science Program (completed)

• data-science-projects — classification, regression, statistical analysis, EDA, and SQL projects from the program.
• gold-ore-purification-ml — ML models predicting gold recovery efficiency across industrial processing stages.
• chicago-taxi-sql-analysis — SQL analysis of Chicago ride data.
• Used Car Price Explorer — a live Streamlit web app analyzing US vehicle listings and price drivers.