IT & Technology / Data Scientist

Turning raw signals into clear decisions.

I design analytical systems that make complex data useful: robust pipelines, interpretable models, thoughtful experiments and visual stories that help teams act with confidence.

Explore selected work Start a conversation
Statistical thinking - ML - experimentation - decision science
38 Models shipped From forecasting systems to classification and ranking.
12 Experiments Designed with measurable hypotheses and decision criteria.
4 Domains Product, finance, operations and customer analytics.
27% Uplift Representative improvement from an optimization program.
01 / Data practice

A matrix mindset.

Useful intelligence grows from connected signals. I combine statistical rigor with product context so analysis survives contact with real decisions.

PythonSQLMachine LearningExperiment DesignForecastingStorytelling
Field note
The best model is the one that changes the quality of the next decision.
02 / Expertise

Tools for finding signal.

Statistical modeling
Machine learning
Data engineering
Experimentation
03 / Working method

From question to signal.

01
Frame

Turn the business question into a measurable analytical objective.

02
Explore

Audit data quality, distributions, leakage, bias and useful relationships.

03
Model

Build a baseline, test alternatives and optimize for the real decision cost.

04
Explain

Package findings into clear recommendations, metrics and next actions.

04 / Selected work

Three projects. Three kinds of signal.

01 / FORECASTING

Demand Pulse

Built a probabilistic demand forecasting workflow that combined seasonality, promotions and operational constraints to improve planning confidence.

PythonTime SeriesForecasting
02 / EXPERIMENTATION

Conversion Lab

Created an experimentation framework for product teams, clarifying hypotheses, guardrails and statistical readouts across multiple launches.

SQLA/B TestingCausal Thinking
03 / ML SYSTEMS

Churn Lens

Developed an interpretable customer-risk model with feature monitoring and decision thresholds designed for practical retention workflows.

ClassificationExplainabilityMonitoring
05 / Contact

Have a hard question? Let's model it.

For data strategy, product analytics, machine-learning projects or experimentation partnerships, send a note.