Summary
PhD Data Scientist with a decade of experience spanning experimental physics, industrial ML, and LLM-powered data pipelines. Specialist in high-dimensional vector modeling, deep neural networks, and scalable NLP systems for complex structured and unstructured datasets.
Work Experience
Staff Data Scientist — Moat Metrics, Spokane, WA (Jun 2024 – Mar 2026)
- Developed a custom deep learning architecture in PyTorch to predict patent entity metrics, with automated feature scaling, LeakyReLU activations, and inference clamping
- Automated patent portfolio valuation scoring at scale using GPU/CPU-agnostic inference and Polars-based data retrieval, deployed on AWS Batch + EC2
- Engineered LLM pipelines for large-scale patent text processing using few-shot and chained prompting, structured output extraction, and local LLM inference via vLLM with KV caching; deployed batch workloads on GCP
- Built an ETL pipeline linking patents to government contracts in Snowflake, resulting in a $250,000 external contract
- Led a data analysis engagement for a private equity firm generating $200,000 in revenue above the base contract
- Contributed to a knowledge graph project spanning 5 million active US patents using batch LLM parsing and Boolean satisfiability algorithms
Research Data Scientist — Aon PLC, Spokane, WA (Sep 2021 – Jun 2024)
- Built an Adaptive Clustering System with a human-in-the-loop workflow, jointly optimizing a PyTorch deep denoising autoencoder and multi-class classifier; user corrections backpropagated through the latent space to align embeddings with expert domain knowledge
- Developed an entity alignment system using Elasticsearch blocking and Conditional Random Fields to resolve patent entity identity across heterogeneous data sources
- Implemented graph traversal algorithms to reconstruct patent assignment histories across complex ownership chains
Senior Research Associate (Postdoctoral) — Duke University, Durham, NC (Jun 2019 – May 2021)
- Characterized trapped ions via microwave and optical transitions; optimized PID laser stabilization and performed interferometric vibration analysis
- Mentored 3 graduate students across physics and ECE departments
Senior Research Associate (Postdoctoral) — University of Oregon, Eugene, OR (Jun 2018 – Jun 2019)
- Conceived and modeled a quantum network based on diamond defect centers; built a scanning confocal microscope for defect measurement
Research Assistant (PhD Candidate) — University of Oregon, Eugene, OR (Jun 2010 – Jun 2018)
- Built Python simulation libraries for light-matter interactions and designed automated data acquisition systems for real-time experiment monitoring
- Directed 4 major projects end-to-end; contributed to 17 peer-reviewed publications and 2 patents
Education
PhD in Physics — University of Oregon, Eugene, OR (2018)
BS in Physics — Washington State University, Pullman, WA (2009)
Skills
Programming & Tools: Python, SQL, PyTorch, Polars, NumPy, Matplotlib, Scikit-Learn, Statsmodels, DuckDB, Postgres, Snowflake, Elasticsearch, Git, Docker, AWS (EC2, S3, Batch), GCP, vLLM, GitLab CI/CD, Terraform
Data Science & ML: Deep Neural Networks, LLM Prompt Engineering (zero-shot, few-shot, chaining), Vector Search and Embeddings, Approximate Nearest Neighbors, Signal Processing, Time-series Analysis, Regression, KMeans and HDBSCAN Clustering, Conditional Random Fields, Human-in-the-Loop, Active Learning, Metric Learning, Manifold Learning, Bayesian Inference
Awards & Honors
- Marthe E. Smith Memorial Science Scholarship (2012)
- Weiser Sr Teaching Assistant Award (2016)
- Optical Society QOST Technical Group Outstanding Poster Presentation (2017)