Turn signals
into decisions.
I build statistical models, experiments and decision systems that move from messy observations to useful action, with an obsession for clarity, reproducibility and measurable outcomes.
Curious by
default.
My practice sits between rigorous statistics and practical product thinking. I frame ambiguous questions, build trustworthy datasets, test hypotheses, communicate uncertainty and turn model output into decisions people can actually use.
Read the
current.
Instead of decorative proficiency bars, expertise is expressed as a model stack: each layer connects a technical method to the decision it helps unlock.
Find the shape inside the noise.
Careful transformations, leakage checks and domain-aware features create models that generalize beyond the notebook.
Measure what changed.
Controlled tests, segmentation and uncertainty intervals keep better tied to evidence rather than intuition.
Predictive modeling
Regression, classification, tree ensembles and calibration.
Causal & experimental thinking
Hypothesis design, A/B testing, uplift and uncertainty.
Data storytelling
Decision-ready dashboards, narratives and executive communication.
From question
to impact.
Senior Data Scientist · Product Intelligence
Lead experimentation and predictive modeling for product teams, partnering with engineering and design to turn behavioral data into measurable product decisions.
Data Scientist · Growth Analytics
Built customer segmentation, propensity models and experiment frameworks that improved prioritization across acquisition and retention programs.
Analytics Associate · Decision Science
Developed reporting systems, statistical analyses and reproducible data workflows for cross-functional business questions.
Models with
a reason to exist.
Retention propensity engine
Combined behavioral signals and calibrated probability scores to help lifecycle teams focus outreach on customers with the strongest intervention potential.
Experimentation command center
Designed an experiment measurement layer that unified cohorts, guardrails and confidence intervals into one decision workflow.
Demand forecasting system
Built a rolling forecast pipeline with seasonal features and scenario bands to support planning under uncertain demand.
Customer signal atlas
Mapped high-dimensional customer behavior into interpretable segments, creating a shared language for product and commercial teams.
Let's test
the hypothesis.
Have a messy dataset, an ambiguous product question or a model that needs a sharper path to impact? Send the context and I'll get back to you.
Typical response: within 2 business days