📊 Digital Finance Forecasting & Analytics Platform
This project is an end-to-end financial analytics platform designed to replicate a modern corporate Finance and FP&A analytical environment.
The platform integrates financial modeling, data engineering, statistical forecasting, and business intelligence into a unified architecture capable of transforming raw financial data into decision-ready analytics and executive insights.
It demonstrates how modern finance teams can move from traditional spreadsheet-based reporting to a scalable analytics-driven finance data stack.
Finance Analytics, Financial Forecasting, FP&A, Financial Modeling, Data Warehouse, Time Series Forecasting, SARIMA Models, Monte Carlo Simulation, Business Intelligence, Power BI, Finance Data Stack.
📌 Project Overview
The project simulates the financial operations of a synthetic industrial engine manufacturer operating across multiple products and sales channels.
The platform integrates several layers of financial analytics:
Financial Modeling → Data Warehouse → Python Automation → Forecasting → BI Dashboards → Streamlit Application
The objective is to demonstrate how financial data can be transformed into structured analytical pipelines capable of supporting forecasting, planning, and executive decision-making.
The platform is structured as a modular financial analytics pipeline, transforming operational financial drivers into decision-ready analytics through an integrated data and analytics stack.
The architecture integrates financial modeling, data warehousing, analytics automation, statistical forecasting, and business intelligence into a unified Finance Data Stack used in modern FP&A environments.
Technology Stack
| Layer | Technology |
|---|---|
| Financial Modeling | Excel (Integrated IFRS 3-Statement Model) |
| Database | PostgreSQL |
| Data Modeling | Star Schema Financial Data Warehouse |
| SQL Development | PostgreSQL (DDL, Views) – managed via DBeaver |
| Data Engineering | Python (Pandas, NumPy, SQLAlchemy) |
| Statistical Modeling | Python (statsmodels, scikit-learn) |
| Forecasting | Time-Series Analysis, SARIMA Models |
| Simulation | Monte Carlo Simulations |
| Visualization | Power BI, Streamlit, Plotly, Matplotlib, Seaborn |
| Development Environment | Jupyter Notebooks, VS Code, Streamlit |
📂 Repository Structure
digital-finance-forecasting-platform
├── module-1-financial-model
│ ├── git_Module1_3statement_IFRS.xlsx
│ └── README.md
│
├── module-2-data-warehouse
│ ├── sql/
│ │ ├── ddl_tables.sql
│ │ ├── views_ifrs.sql
│ │ └── load_data.sql
│ ├── data/
│ ├── assets/
│ │ └── star_schema_diagram.png
│ └── README.md
│
├── module-3-financial-reporting-automation
│ ├── Module3.1_py_anomaly.ipynb
│ ├── Module3.2_reporting.ipynb
│ └── README.md
│
├── module-4-forecasting
│ ├── Module4.1_py_reg.ipynb
│ ├── Module4.2_py.ipynb
│ ├── Module4.3_excel.pdf
│ ├── Module4.4_py_anomaly.ipynb
│ ├── Module4.5_fpa_bridge.ipynb
│ ├── Module4.6_ml_forecast.ipynb
│ └── README.md
│
├── module-5-bi-dashboard
│ ├── powerbi/
│ │ └── Power_BI_finance_ifrs_dw.pbix
│ ├── assets/
│ │ └── dashboard_preview.pdf
│ └── README.md
│
├── module-6-streamlit-dashboard
│ ├── app.py
│ ├── data
│ └── README.md
│
└── README.md
📊 Business Context
The platform models the financial evolution of a synthetic industrial engine manufacturer operating through:
Sales Channels:
- Direct Sales
- Retail Distribution
- Online Sales
Product Lines:
- Automotive Engines
- Industrial Engines
- Electric Solutions
The financial dataset spans 2010–2024, capturing a full corporate cycle:
- Growth phase
- Operational stress period (2018–2019)
- Post-crisis recovery (2020–2024)
This structure allows the simulation of realistic financial dynamics such as revenue growth, margin compression during stress periods, and financial stabilization during recovery.
📦 Project Modules
Development of an integrated IFRS-compliant 3-Statement financial model.
Includes:
- Income Statement
- Balance Sheet
- Cash Flow Statement
- Financial driver assumptions
- Revenue seasonality allocation
- Monthly financial dataset generation
The model serves as the financial engine of the platform.
Implementation of a PostgreSQL Financial Data Warehouse using a Star Schema architecture.
The warehouse exposes IFRS reporting views used for analytics and financial reporting.
A detailed explanation of the dimensional model, including the structure of dimension tables, fact tables, and the datasets contained in the data/ directory, is available in the Module 2 README. Please refer to that documentation for the complete schema description and data definitions.
Python-based pipeline to extract, validate, and transform financial data from the Data Warehouse, enriched with anomaly detection capabilities.
Key components:
- SQLAlchemy data extraction from PostgreSQL
- Pandas-based transformations and KPI calculations
- data validation and financial integrity checks
- multi-layer anomaly detection (Z-score, STL, Isolation Forest)
- ensemble scoring for robust anomaly identification
- standardized reporting outputs (CSV / Excel)
The module replaces manual reporting with a reproducible financial analytics and monitoring workflow.
Advanced financial analytics layer responsible for transforming historical financial data into forward-looking insight, financial plans, and performance explanations.
The module implements a structured FP&A workflow:
diagnostic analysis → statistical forecasting → financial planning → forecast governance → variance analysis → driver-based forecasting enhancement
Core analytical components include:
- Statistical diagnostics and model selection, including EDA, stationarity testing, seasonality analysis, and feature evaluation
- SARIMA-based forecasting, generating multi-horizon projections with probabilistic confidence intervals
- Scenario modeling, supporting FP&A planning through base, upside, and downside financial projections
- Forecast governance, using anomaly detection frameworks to validate accuracy, plausibility, and consistency
- Variance analysis, decomposing performance into price, volume, and mix effects across business segments
- Driver-based machine learning models, capturing nonlinear relationships between revenue and operational drivers (price, volume, channel, product mix)
The module integrates statistical rigor with business interpretability, enabling a complete FP&A intelligence cycle — from forecast generation to validation and performance explanation — supporting data-driven financial decision-making.
Power BI dashboards transform the analytical outputs into interactive executive insights.
The dashboard includes three analytical perspectives:
- Financial Performance Journey
- Revenue Drivers Analysis
- Forecast & Scenario Planning
The goal is to translate complex financial analytics into decision-support dashboards.
📈 Key Analytical Capabilities
The platform demonstrates several advanced finance analytics capabilities:
- Integrated IFRS financial modeling
- Financial data warehouse design
- automated financial reporting pipelines
- statistical forecasting models
- probabilistic scenario analysis
- executive financial dashboards
Development of an interactive Streamlit application for financial analysis and executive exploration.
Includes:
- Financial Journey view
- Revenue Drivers analysis
- Forecast & Scenario Analysis
- dynamic filters and KPI cards
- PostgreSQL integration via SQLAlchemy
- interactive visualizations with Plotly
The module transforms the platform into a web-based analytical application for interactive decision support.
The Digital Finance Forecasting & Analytics Platform demonstrates how finance teams can transform fragmented financial data into a structured analytics architecture capable of supporting forecasting, planning, and executive decision-making.
📌 Financial Data Integration & Governance
Corporate finance environments often rely on disconnected spreadsheets and manual reporting processes. This platform demonstrates how financial data can be centralized in a governed Financial Data Warehouse structured around IFRS reporting logic.
📌 Automated Financial Reporting
The Python automation pipeline replaces manual reporting workflows with reproducible data processes and automated KPI generation.
📌 Data-Driven Forecasting & Risk Awareness
Statistical forecasting models (SARIMA) and Monte Carlo simulations allow finance teams to quantify revenue uncertainty and evaluate financial scenarios.
📌 FP&A Planning & Scenario Analysis
Forecast outputs are integrated into an FP&A planning framework supporting budgeting, scenario evaluation, and variance analysis.
📌 Executive Decision Support
Power BI dashboards translate complex financial analytics into intuitive executive insights for strategic decision-making.
⚠ Data Disclaimer
The dataset used in this project represents synthetic financial data created to simulate realistic corporate financial dynamics.
The purpose is to demonstrate financial analytics architecture and methodology, not to represent a real company dataset.
🔧 Environment Setup
Install required Python dependencies:
pip install -r requirements.txt
Configure database credentials using .env.example:
DB_USER=your_user
DB_PASS=your_password
DB_HOST=localhost
DB_PORT=5432
DB_NAME=finance_ifrs_dw
📬 Contact
📩 Email: gustavo.provento@gmail.com
💼 LinkedIn: linkedin.com/in/gustavo-m-freitas
📂 GitHub: github.com/gustavo-m-freitas
