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.

Each project ships with a live walkthrough (futurist dark page, hosted on GitHub Pages) and a full repo with code, business memo, and reproducibility instructions.

• dbt Analytics Warehouse · 🌐 walkthrough — Production-shaped dbt project on DuckDB. 5 sources → 5 staging → 3 intermediate → 6 marts → 1 snapshot, 70 tests, 2 exposures. Portable to Snowflake/BigQuery with a profile change. dbt · DuckDB · SQL · Python

• A/B Testing Framework · 🌐 walkthrough — End-to-end experimentation toolkit on a 60k-user synthetic experiment: power analysis, SRM (chi²), CUPED, Benjamini-Hochberg, auto-generated decision doc. Headline: CVR +9.4% (p=0.0068), revenue +$0.225/visitor (p=0.0097), 0/8 segments significant after BH → ~$984k annualized uplift. Python · statsmodels · scipy

• SaaS MRR, Churn & Cohort Analytics · 🌐 walkthrough — Subscription analytics on a synthetic SaaS dataset (5,000 customers · 36 months · 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×. Churn concentrates in months 3–6. Python · pandas · matplotlib

• Olist E-commerce SQL Analytics · 🌐 walkthrough — End-to-end SQL on Olist-shaped Brazilian e-commerce (96,497 orders, 92,315 customers, R$ 14.7M revenue). DuckDB warehouse, Kimball star, 10 business questions. Headline: late deliveries collapse review score to 1.64/5 vs 3.98; bottleneck is carrier transit (25.6d) not seller dispatch (8d); R$ 560k revenue at risk. SQL · DuckDB · Python

• NYC 311 Operational Dashboard · 🌐 walkthrough — Operational analytics on live NYC Open Data (100k 311 requests via Socrata API). DuckDB warehouse, 8 charts, interactive Plotly map, 4-page Power BI wireframe with DAX. Headline: 7.5pp equity gap between Staten Island (85.0% SLA) and the Bronx (77.5%) — slower and worse in the same borough. Python · requests · DuckDB · Plotly · Power BI

• RFM Customer Segmentation · 🌐 walkthrough — Customer segmentation on UCI Online Retail II (805,549 transactions, 5,878 customers, £17.7M revenue). RFM quintile scoring + 11 canonical segments + k-means parity + per-segment marketing playbook. Headline: 49.5% of revenue from 609 Champions; quarterly playbook $13.7k cost → $144.7k projected = 10.5× blended ROI. Python · scikit-learn · DuckDB · SQL

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

Adaptive section generated from GitHub repo metadata. Last update: 2026-05-30 10:54 UTC.

• Public/source repos tracked: 12
• Detected languages: 7
• Most recent repo update: 2026-05-26

DuckDB · Plotly · Power BI · dbt · pandas · scikit-learn

• ednarzinho — Biblioteca Python didatica de matrizes: determinante, inversa e geracao de notebook .ipynb. pip install ednarzinho Python · ⭐ 0 · updated 2026-05-26
• estudo-de-caso-testes — No description yet. Python · ⭐ 0 · updated 2026-05-24
• bigquery-airflow-dbt-warehouse — Cloud-native analytics warehouse — dbt on BigQuery thelook_ecommerce + Airflow (Astronomer Cosmos) orchestration + GitHub Actions CI. Kimball star schema, 21 models, 88… Python · ⭐ 0 · updated 2026-05-24 · airflow · analytics-engineering · astronomer-cosmos · bigquery
• marketing-funnel-streamlit — Multi-touch attribution + funnel + CAC/ROAS Streamlit dashboard. Synthetic 75k users / 1,143 paid / $172k spend / 5 channels — finds paid_social structurally unprofitabl… Python · ⭐ 0 · updated 2026-05-23 · attribution · cac · dashboard · data-visualization
• rfm-customer-segmentation · walkthrough — Customer segmentation on the UCI Online Retail II dataset (805,549 transactions, 5,878 customers, £17.7M revenue): RFM quintile scoring, 11-segment rule-based classifica… Python · ⭐ 0 · updated 2026-05-22 · clustering · customer-segmentation · duckdb · ecommerce-analytics
• nyc-311-dashboard · walkthrough — Operational analytics on NYC 311 service requests — Python + DuckDB pipeline, SLA + equity analysis across 5 boroughs, Power BI dashboard spec. HTML · ⭐ 0 · updated 2026-05-22 · dashboard · data-analysis · duckdb · geospatial

Ranked by recent activity, repo completeness, homepage/walkthrough presence, topics, stars, and language diversity — not by a manually fixed list.

• nyc-311-dashboard · walkthrough — Operational analytics on NYC 311 service requests — Python + DuckDB pipeline, SLA + equity analysis across 5 boroughs, Power BI dashboard spec. HTML · ⭐ 0 · updated 2026-05-22 · dashboard · data-analysis · duckdb · geospatial
• olist-ecommerce-analytics · walkthrough — End-to-end SQL analytics on a synthetic Olist e-commerce dataset — DuckDB warehouse, Kimball star schema, 10 business-question queries, Python chart deck. Python · ⭐ 0 · updated 2026-05-22 · analytics-engineering · data-analytics · duckdb · ecommerce
• dbt-analytics-warehouse · walkthrough — Production-shaped dbt project on DuckDB - synthetic e-commerce + SaaS data modeled into a Kimball warehouse with 70 tests, snapshots, exposures, and reusable macros. Python · ⭐ 0 · updated 2026-05-22 · analytics-engineering · data-engineering · data-modeling · data-quality
• rfm-customer-segmentation · walkthrough — Customer segmentation on the UCI Online Retail II dataset (805,549 transactions, 5,878 customers, £17.7M revenue): RFM quintile scoring, 11-segment rule-based classifica… Python · ⭐ 0 · updated 2026-05-22 · clustering · customer-segmentation · duckdb · ecommerce-analytics
• ab-testing-framework · walkthrough — End-to-end A/B testing framework — power analysis, SRM checks, CUPED variance reduction, BH-corrected segment breakouts, and an auto-generated ship/no-ship decision doc.… Python · ⭐ 0 · updated 2026-05-22 · ab-testing · bootstrap · causal-inference · cuped
• saas-mrr-churn-analytics · walkthrough — Subscription analytics on a synthetic SaaS dataset — 5,000 customers, 36 monthly billing snapshots, MRR waterfall, NRR/GRR cohorts, Kaplan-Meier survival, LTV/CAC by tie… Python · ⭐ 0 · updated 2026-05-22 · churn-analysis · cohort-analysis · data-analytics · kaplan-meier

_{This block is regenerated by GitHub Actions. New Python, JavaScript, TypeScript, SQL, notebook, or web projects will appear automatically after the next run.}