ETL → Star Schema → KPIs → Dashboards (Power BI) with BI governance (read-only role & contract views).
Portfolio project simulating an enterprise-grade financial analytics pipeline.

⬇️ Download PBIT template · 📺 See dashboard pages · 📘 Detailed README

• Where does revenue come from (acquisition vs. returning) and which segments/categories drive margin?
• Are sales improving MoM and YoY? Which quarters/months concentrate demand?
• Which ship mode sells the most and at what logistics burden? Are we meeting the ≤ 4-day SLA?
• Is there backlog (orders vs. shipments)?

• 📂 sql_scripts/ → SQL scripts organized by ETL, Modeling, BI, and Admin.
• 📂 docs/ → Documentation, diagrams, and design notes.
• 📂 dashboards/ → Power BI (.pbix/.pbit), screenshots, and visual themes.

• 🔹 ETL → See scripts
• 🔹 Modeling → See scripts
• 🔹 BI → See scripts
• 🔹 Admin → See scripts
• 🔹 Docs → See documentation
• 🔹 Dashboards → See dashboards

This repository demonstrates a full Financial Analytics workflow:

1. ETL → Load and cleanse raw data (STAGE → RAW → CLEAN).
2. Modeling → Star schema (FACT + DIM) and financial KPIs.
3. BI → Business views for Power BI dashboards.
4. Admin → Security, performance, and governance.
5. Docs → ER diagrams and design decisions.
6. Dashboards → Final storytelling with Power BI.

Revenue mix, segment performance & margin, discount vs. margin with thresholds.

Key insights

• Revenue mix: ≈ 70% of revenue comes from returning customers consistently (2011–2014).
• Segment performance: Consumer leads sales (≈ $6.5M); Home Office shows the highest margin (≈ 11.9%).
• Category breakdown: Technology leads sales ($4.7M, margin ≈ 14%); Furniture sells similarly ($4.1M) with lower margin (≈ 6.9%).
• Discount vs. margin: medians margin 13.8% and discount 9%; Tables falls in high discount / negative margin quadrant (avoid).

MoM/YoY trends, best quarters/months; heatmap by month/year.

Key insights:

• Crecimiento YoY sostenido (panel muestra ~47%), con aceleraciones entre ago–nov.
• Q4 es el trimestre pico cada año (2014 Q4 ~$1.49M).
• Diciembre domina el mes pico de ventas de forma consistente.

SLA (≤4 días) por modo, ventas vs. shipping %, órdenes vs. despachos (backlog).

Key insights

• SLA (≤ 4 days): Same Day (0d), First (2d) and Second (3d) meet target; Standard = 5d → out of SLA.
• Efficiency by mode: Standard drives sales ($7.6M) with lower shipping burden (8.1%) and $40.61/order. Same Day is the most expensive (≈17.2%, $86/order).
• Orders vs. shipments: Year-end shipments > orders indicates backlog clearance; January rebalances.

• Safe deltas: robust Safe % vs LM/LY against slicers and missing months.
• Time intelligence: YTD, MoM, YoY measures.
• Usability: context-aware tooltips, KPI labels, curated views for BI consumption.

• Read-only BI user (bi_reader) with SELECT-only privileges.
• Published contract views under the global_superstore_bi schema.

1) SQL

• Create the DB and run scripts in sql_scripts in order: etl/ → modeling/ → bi/ → admin/.
• The read-only BI user bi_reader (role & grants) is created in sql_scripts/admin.

2) Power BI

• Open the template: dashboards/powerbi/templates/GlobalSuperstore_Finance_Dashboard.pbit.
• Point the connection to schema global_superstore_bi (contract views).
• Refresh the model.

Need the full step-by-step? See Detailed README.

• Source: Global Superstore Dataset
• Use: Public dataset for Data Analytics practice.

• SQL (MySQL / compatible)
• Power BI (DAX)
• GitHub (documentation & version control)

Project by Daiana Beltrán
LinkedIn · GitHub