I don’t just extract data — I model systems.
I combine business strategy, Python engineering, and applied mathematics
to improve operational decisions in high-stakes environments.
Data without structure is noise. I use modeling, statistics, and logic to isolate signal under uncertainty.
Engineering is a means. Every model and dashboard is tied to decisions: efficiency, risk, and resource allocation.
Python ETL, SQL, automation, and clean interfaces — built for maintainability, auditability, and scale.
Open-source tooling and case studies (with confidentiality preserved).
Classifies time-clock punches as IN / OUT / ERROR and supports shift inference in 24/7 rotating schedules. Public version uses a synthetic dataset that preserves real-world complexity.
A personal Python library that standardizes ETL patterns and reporting utilities for reproducible analytics work.
Designed a risk scoring approach to detect high-friction operational events. Focus: robust evaluation, bias checks, and explainability under policy constraints.
Built a parsing pipeline to convert free-text legacy logs into structured tables for analytics and governance.
Short write-ups on modeling, algorithms, and how math becomes operational decisions.
How to model sequences of punches, engineer features, avoid leakage, and evaluate by employee groups.
Read → Applied CalculusUsing marginal analysis to detect diminishing returns and identify decision-ready inflection points.
Read →