Financial Analytics Portfolio

Single Source of Truth (SSOT) — Master plan for building a comprehensive financial data analyst portfolio.

---

Portfolio Overview

Purpose

Develop a professional portfolio demonstrating expertise in:

• Financial modeling & forecasting
• AP/AR automation
• Business intelligence & dashboarding
• Data cleaning & validation
• Applied AI/ML workflows
• Python/R analytics
• SQL data modeling
• Operational finance

Target Outcome

4-6 deep, end-to-end projects that prove technical competence and business acumen to recruiters and hiring managers.

---

Project List

#	Project	Status	Description
1	Pizza Sales Revenue Forecasting	Complete	Time series forecasting using R (ARIMA)
2	Invoice Anomaly Detection	Complete	Python anomaly detector for AP transactions (duplicates, outliers, weekend/holiday flags)
3	Executive Dashboard	Complete	Python KPI dashboard with revenue trends, profitability analysis
4	Financial Modeling	Complete	Python 3-statement model with Base/Upside/Downside scenarios
5	D.A.T.A.	Complete	Dashboard & Analytics Tool for Accounting — Streamlit app deployed on Railway

---

Project Template (Standard Structure)

Each project follows this structure:

project-XX-name/
├── data/
│   ├── raw/           # Original datasets
│   └── processed/     # Cleaned/transformed data
├── notebooks/         # Jupyter/R notebooks
├── src/               # Python/R scripts
├── dashboards/        # Tableau workbooks, screenshots
├── reports/           # Final writeups, PDFs
└── README.md          # Project-specific documentation

Required Sections in Each Project README

1. Problem Statement — Clear business problem and why it matters
2. Dataset Overview — Source, contents, columns, known issues
3. Tools & Skills — Technologies and financial concepts demonstrated
4. Data Cleaning & Prep — Missing values, validation, feature engineering
5. Analysis/Model/Automation — Technical workflow details
6. Visualizations — Charts, dashboards, KPI summaries
7. Business Interpretation — Translated into decision-making language
8. Deliverables — Links to notebooks, dashboards, apps
9. Future Enhancements — Potential improvements

---

Execution Plan

Phase 1: Setup

• Create folder structure
• Initialize GitHub repository
• Document dependencies
• Set up Python/R environments

Phase 2: Project 1 — Pizza Sales Revenue Forecasting

• Download Kaggle pizza sales dataset
• Exploratory data analysis (6 business questions)
• Build forecasting model (ARIMA)
• Evaluate model accuracy
• Create visualizations
• Write project README
• Generate final report (HTML + PDF)
• Document lessons learned (Christmas seasonality)

Phase 3: Project 2 — Invoice Anomaly Detection

• Find AP/invoice transaction dataset (50K records)
• Build Python anomaly detection script
• Create anomaly detection rules:
  • Duplicate invoices (same vendor + amount within 7 days)
  • Statistical outliers (>3 std dev from vendor mean)
  • Round number amounts (fraud indicator)
  • Weekend/holiday invoice dates
  • New vendor spikes (first invoice > $5K)
  • Inactive vendor reactivation (>180 days gap)
• Generate HTML report and CSV outputs
• Document methodology (python_vs_excel_summary.txt)

Phase 4: Project 3 — Executive Dashboard

• Download Superstore dataset from Kaggle
• Prep data with Python (calculated fields, margins)
• Build Python KPI dashboard:
  • KPI tiles (Revenue, Profit, Margin, Orders)
  • Revenue trend over time
  • Profit by category breakdown
  • Top products and regions analysis
• Generate interactive HTML report

Phase 5: Project 4 — Financial Modeling

• Create historical restaurant financials (24 months)
• Build Python 3-statement model:
  • Income Statement
  • Balance Sheet
  • Cash Flow Statement
• Implement 3 scenarios (Base/Upside/Downside)
• Create 5-year projections (2024-2028)
• Document assumptions (assumptions.md)
• Generate formatted Excel workbook with Executive Summary
• Add sensitivity analysis

Phase 6: Project 5 — D.A.T.A. (Dashboard & Analytics Tool for Accounting)

• Design app architecture
• Build Streamlit interface with password protection
• Implement smart column detection (dates, amounts, categories, entities)
• Add data cleaning pipeline (currencies, dates, missing values)
• Create KPI dashboard with real-time calculations
• Add interactive Plotly visualizations (trends, pie charts, bar charts)
• Deploy to Railway (live at https://finance-copilot-production-38c7.up.railway.app)
• Create landing page with demo password for hiring managers

Phase 7: Portfolio Website (GitHub Pages)

• Configure GitHub Pages with Jekyll
• Apply Leap Day theme for professional appearance
• Create index.html landing pages for each project
• Ensure all reports visible directly on project pages
• Connect custom domain (optional)

---

Technical Dependencies

Python

pandas>=2.0.0
numpy>=1.24.0
matplotlib>=3.7.0
seaborn>=0.12.0
plotly>=5.14.0
scikit-learn>=1.2.0
statsmodels>=0.14.0
prophet>=1.1.0
streamlit>=1.22.0
pytesseract>=0.3.10
opencv-python>=4.7.0
python-dotenv>=1.0.0

R (for forecasting)

tidyverse
lubridate
forecast
readxl
scales
knitr
kableExtra

Tools

• Python 3.11+
• R 4.3+
• RStudio
• VS Code / Cursor
• Tableau Public or Power BI
• Figma (design)
• Railway (deployment)
• GitHub (version control)

---

Portfolio Website

Live Site

GitHub Pages: https://jkatz015.github.io/portfolio

Structure

1. Home — README renders as main landing page
2. Project Pages — Each project has its own index.html with interactive reports
3. Live Demo — D.A.T.A. app hosted on Railway with demo access

Design

• Leap Day Jekyll theme for professional appearance
• Light theme across all reports (#f5f5f5 background, #4CAF50 green accents)
• Mobile responsive
• Fast loading static HTML

Tech Stack

• Theme: Jekyll Leap Day
• Reports: Python-generated HTML with Plotly
• App Hosting: Railway (Streamlit)
• Version Control: GitHub

---

Progress Tracker

Date	Milestone	Notes
Nov 2024	Repository created	Initial setup
Nov 2024	Project 1 complete	Pizza sales forecasting with R
Dec 2024	Project 2 complete	Python AP anomaly detector (50K records, 7 detection rules)
Dec 2024	Project 3 complete	Python KPI executive dashboard (revenue trends, profitability)
Dec 2024	Project 4 complete	Python 3-statement financial model with scenarios
Dec 2024	Project 5 complete	D.A.T.A. Streamlit app deployed to Railway
Dec 2024	GitHub Pages live	Leap Day theme, all projects with landing pages

---

Resources

Datasets

• Kaggle Pizza Sales — Project 1
• Superstore Dataset
• Synthetic Invoice Data (TBD)

Learning

• R for Data Science
• Forecast Package Documentation
• Streamlit Documentation
• Tableau Public

---

License

This portfolio is for educational and professional demonstration purposes.

---

Last updated: December 2024