🤖 AI-Driven ESG Financial Analytics Platform

An end-to-end financial analytics platform simulating real-world investment banking workflows — integrating SQL, Python, Excel and Power BI with AI-driven ESG scoring, NLP sentiment analysis and ML-based ESG forecasting to generate investment recommendations across a 10-company S&P 500 coverage universe.

---

🏗️ System Architecture

Data Sources (Yahoo Finance, Kaggle ESG, Simulated News)
         ↓
Layer 1: SQL Server Data Warehouse (5 tables, 7,100+ rows)
         ↓
Layer 2: Python Analytics Engine (9 modules)
         ↓
    ┌────┴────┐
    ↓         ↓
Layer 3:   Layer 4:
Excel      Power BI
DCF Model  Dashboard
    ↓         ↓
    └────┬────┘
         ↓
Layer 5: AI & ESG Enhancement
(Sentiment + ML Prediction + Recommendation Engine)

---

📊 Coverage Universe

Ticker	Company	Sector	ESG Rating
MSFT	Microsoft Corporation	Technology	AAA
AAPL	Apple Inc.	Technology	AAA
NEE	NextEra Energy	Utilities	AAA
UL	Unilever PLC	Consumer Staples	AA
JNJ	Johnson & Johnson	Healthcare	AA
JPM	JPMorgan Chase	Financials	AA
AMZN	Amazon.com Inc.	Consumer Discretionary	A
TSLA	Tesla Inc.	Consumer Discretionary	A
CVX	Chevron Corporation	Energy	BBB
XOM	Exxon Mobil	Energy	BBB

---

🗂️ Project Structure

esg-financial-analytics-platform/
│
├── python/
│   ├── db_connection.py          ← SQL Server connector
│   ├── ingest_stock_prices.py    ← yfinance data pipeline
│   ├── financial_metrics.py      ← Ratio calculations
│   ├── esg_scoring.py            ← ESG composite scoring
│   ├── sentiment_analysis.py     ← NLTK VADER NLP
│   ├── esg_predictor.py          ← ML ESG forecasting
│   ├── ai_recommendation.py      ← AI scoring engine
│   ├── export_master.py          ← Data export pipeline
│   └── pipeline_validation.py    ← End-to-end validator
│
├── sql/
│   ├── schema.sql                ← Database schema & tables
│   └── queries.sql               ← Analytical queries
│
├── data/processed/
│   ├── master_dataset.csv
│   ├── financial_metrics.csv
│   ├── esg_scores_processed.csv
│   ├── esg_investment_tiers.csv
│   ├── esg_predictions_full.csv
│   ├── sentiment_scores.csv
│   ├── esg_predictions_2024.csv
│   └── ai_recommendations.csv
│
├── excel/
│   └── financial_model.xlsx      ← 7-sheet DCF model (Apple Inc.)
│
├── powerbi/
│   └── ESG_Dashboard.pbix        ← 5-page interactive dashboard
│
├── docs/
│   ├── architecture_diagram.svg
│   └── ESG_Project_Summary.docx
│
└── requirements.txt

---

⚙️ Key Features

🗄️ SQL Data Warehouse

• 5 normalized tables with foreign key relationships
• 7,100+ rows across companies, financials, stock prices, ESG scores and news sentiment
• T-SQL analytical queries for ESG vs financial performance

🐍 Python Analytics Engine (9 Modules)

• Automated stock price ingestion via yfinance API
• Financial ratio computation (margins, CAGR, ROE)
• Weighted ESG composite scoring — Environmental 40%, Social 30%, Governance 30%
• ESG-adjusted WACC framework (±75bps based on ESG rating)

📊 Excel Financial Model (Apple Inc.)

• Full 3-statement model (FY2021–FY2026E)
• DCF valuation with ESG-adjusted WACC
• Dynamic Bear / Base / Bull scenario toggle
• Sensitivity table (WACC × Terminal Growth Rate)
• Peer comparison across all 10 companies

📈 Power BI Dashboard (5 Pages)

• Executive Overview with KPI cards
• Financial Performance analysis
• ESG Deep Dive with sub-score breakdown
• Investment Recommendations with tier coloring
• AI Insights with sentiment and predictions

🤖 AI & ESG Integration

• NLTK VADER sentiment analysis across 60 headlines
• Scikit-learn Linear Regression ESG forecasting
• 3-pillar AI scoring — Financial 35%, ESG 40%, Sentiment 25%
• Confidence-rated investment recommendations

---

🎯 AI Recommendation Results (FY2023)

Ticker	AI Score	Recommendation	Confidence
MSFT	75.22	✅ Strong Buy	Medium
AAPL	74.25	✅ Strong Buy	Medium
NEE	61.21	🟢 Buy	Low
JPM	53.03	🟡 Hold	Low
UL	48.86	🟡 Hold	Low
JNJ	46.24	🟡 Hold	Low
AMZN	42.52	🟠 Underperform	Low
TSLA	32.70	🟠 Underperform	Medium
CVX	28.65	🔴 Sell	Medium
XOM	18.55	🔴 Sell	High

---

💡 Key Analytical Insights

1. ESG Premium — AAA-rated companies (MSFT, AAPL, NEE) receive a 75bps WACC reduction vs BBB-rated energy majors, directly impacting intrinsic value
2. ESG-Margin Correlation — Technology sector shows the strongest positive correlation between ESG score and net profit margin
3. Sentiment Divergence — CVX shows positive news sentiment (+0.33) despite a poor ESG score (44.0), illustrating why multi-factor models outperform single-metric approaches
4. Energy Transition — NEE (clean energy) shows 43.6% revenue CAGR vs XOM's 20.9%, supporting the ESG-growth thesis

---

🚀 How to Run

Prerequisites

• Python 3.10+
• Microsoft SQL Server + SSMS
• Power BI Desktop (free)
• Microsoft Excel

Setup

# Install dependencies
pip install -r requirements.txt

# Run the full pipeline in order:
python python/db_connection.py
python python/ingest_stock_prices.py
python python/financial_metrics.py
python python/esg_scoring.py
python python/export_master.py
python python/sentiment_analysis.py
python python/esg_predictor.py
python python/ai_recommendation.py
python python/pipeline_validation.py

Database

-- Run in SSMS:
-- 1. sql/schema.sql    (create tables)
-- 2. sql/queries.sql   (run analytical queries)

---

🛠️ Skills Demonstrated

Python SQL Power BI Excel DAX Machine Learning NLP Financial Modelling DCF Valuation ESG Analysis Data Pipeline Investment Research

---

👤 Author

Vansh Goyal

• 📧 [vggoyal2304@gmail.com]
• 💼 [www.linkedin.com/in/vansh-goyal-fin]

---

⚠️ Disclaimer

Built for educational and portfolio purposes. Does not constitute financial advice or a recommendation to buy or sell any security. All data sourced from publicly available sources (Yahoo Finance, Kaggle).