Building data products that inform and scale
Senior Data Scientist with 12 years across healthcare AI, fintech, and product analytics. I build data products from concept to production — RAG assistants, risk models, evaluation frameworks, and clinical analytics tools.
About me
I'm a Senior Data Scientist based in Dallas, TX, with 12 years of experience building data products at PathAI, Tempus AI, Mercury, and Meta.
My work spans healthcare AI, fintech risk, and product analytics. I've built RAG-based tools for clinical and scientific users, designed LLM evaluation frameworks for retrieval relevance, answer faithfulness, and hallucination risk, and developed risk models and experimentation systems that informed product decisions at scale.
When I'm not building data products, I'm thinking about model reliability, evaluation quality, and how to make complex AI systems trustworthy and understandable to the teams that use them. I hold an M.S. in Data Science and a B.S. in Computer Science from The University of Sydney.
Jul 2024–Present
Senior Data Scientist, AI Pathology & LLM Products at PathAI — model reliability analysis, RAG tools, LLM evaluation frameworks.
Oct 2021–Jun 2024
Senior Data Scientist, Real-World Evidence & Clinical AI at Tempus AI — oncology analytics, cohort logic, NLP- and LLM-assisted tools.
Apr 2018–Sep 2021
Data Scientist at Mercury — fraud risk models, growth analytics, internal search and chatbot prototypes.
Jun 2014–Mar 2018
Data Scientist, Product Analytics at Meta — large-scale product analytics, A/B testing, NLP classification.
2012–2014
M.S. Data Science — The University of Sydney.
2008–2012
B.S. Computer Science — The University of Sydney.
How I work
Statistical Rigor
Evidence-based by default. I design experiments with proper controls, apply causal inference, and validate model assumptions before making recommendations.
Healthcare AI Impact
Built RAG assistants and analytics tools that give clinicians and scientists fast, trustworthy access to model outputs, cohort definitions, and validation results.
LLM Evaluation
Designed evaluation frameworks assessing retrieval accuracy, answer faithfulness, hallucination risk, and citation quality — keeping AI products honest and reliable.
Experimentation Culture
Designed and analyzed A/B tests and lifecycle experiments across onboarding, pricing, activation, and product features, from hypothesis and guardrail metrics through launch recommendation.
Data Products
Risk models, dashboards, cohort tools, and evaluation workflows that turn complex clinical, risk, and product data into reliable models and production-ready workflows.
Cross-Team Partnership
Trusted partner to product, engineering, clinical, risk, compliance, and executive teams — translating complex data findings into clear, actionable decisions.