Data Analyst turning raw data into business decisions. SQL · Python · Power BI · BigQuery. I build dashboards, ETL pipelines, and analytics that ship — with a strong interest in experimentation, A/B testing, and analytics engineering on the modern data stack.

• 📈 Analytics & Experimentation: A/B testing, hypothesis testing, cohort & funnel analysis, KPI design.
• 🧱 Modern Data Stack: dbt models, data marts, semantic layers, data quality testing.
• 📊 Dashboards & Reporting: Executive reporting, product analytics, marketing attribution, financial KPIs.
• 🐍 Python for Analytics: pandas, NumPy, matplotlib, seaborn, scikit-learn for predictive modeling.
• 🧮 Statistics: Regression, segmentation, time-series, confidence intervals, lift calculation.
• 🗣️ Data Storytelling: Translating findings into clear narratives for non-technical stakeholders.

• SaaS MRR, Churn & Cohort Analytics — Subscription analytics on a synthetic SaaS dataset (5,000 customers · 36 monthly snapshots · 3 tiers). MRR waterfall, NRR/GRR cohorts, Kaplan-Meier survival, LTV/CAC. Headline: ending MRR $1.99M, median NRR 85.2%, 12-month survival 53.9%, LTV/CAC up to 17.0× (Growth tier). Diagnosed churn concentrating in months 3–6 of customer life (7–9% vs 2–3% early). Python · pandas · matplotlib
• Olist E-commerce SQL Analytics — End-to-end SQL analytics on an Olist-shaped Brazilian e-commerce dataset (96,497 delivered orders, 92,315 customers, R$ 14.7M revenue). DuckDB warehouse, Kimball star schema, 10 business-question queries. Headline: late deliveries collapse review score to 1.64/5 vs 3.98 on-time; 86.8% of 8+d-late orders rated 1–2 stars; bottleneck is carrier transit (25.6d) not seller dispatch (13.1d); R$ 560k revenue-at-risk. SQL · DuckDB · Python
• Portfolio Hub — Index of academic and personal data-science / ML projects from CEUB. New analytics projects shipping daily this week.

"In data we trust — but only after we test it."