Upload 35 files

Dockerfile

@@ -1,13 +1,13 @@
-FROM python:3.11-slim
-
-WORKDIR /app
-
-COPY requirements.txt .
-
-RUN pip install --no-cache-dir -r requirements.txt
-
-COPY . .
-
-EXPOSE 7860
-
-CMD ["uvicorn", "API.main:app", "--host", "0.0.0.0", "--port", "7860"]
+FROM python:3.11-slim
+
+WORKDIR /app
+
+COPY requirements.txt .
+
+RUN pip install --no-cache-dir -r requirements.txt
+
+COPY . .
+
+EXPOSE 7860
+
+CMD ["uvicorn", "API.main:app", "--host", "0.0.0.0", "--port", "7860"]

README.md

@@ -1,10 +1,148 @@
+# 🌱 Intelligent ESG Scoring System for Enterprise Decision-Making
+
+An AI-driven, multi-agent ESG intelligence platform that enables real-time ESG monitoring, predictive risk scoring, and automated regulatory compliance for enterprises.
+
+Built for HackAvensis26 under the Artificial Intelligence & Machine Learning track.
+
+---
+
+## 📌 Project Overview
+
+Enterprises today face major challenges with manual ESG reporting, delayed insights, and rapidly changing regulations. This project introduces a real-time, explainable, and scalable ESG scoring system powered by AI-driven multi-agent architecture.
+
+---
+
+## 🎯 Problem Statement
+
+- ESG compliance is managed through manual and delayed reporting  
+- Corporates lack real-time visibility into ESG risks  
+- Regulatory frameworks are complex and frequently changing  
+- Late detection leads to penalties, reputational damage, and loss of trust  
+
+---
+
+## ✅ Objectives
+
+- Enable real-time ESG monitoring  
+- Provide explainable ESG insights  
+- Predict ESG risks before escalation  
+- Automate regulatory compliance checks  
+
+---
+
+## 💡 Proposed Solution
+
+### AI-Driven Multi-Agent ESG Intelligence Platform
+
+The platform consists of specialized agents, each responsible for a specific ESG function:
+
+- Operational Agent – Monitors operational ESG metrics  
+- Financial Agent – Analyzes financial and investment-related ESG data  
+- Regulatory Agent – Ensures compliance with evolving ESG regulations  
+- Risk Agent – Detects anomalies and predicts future ESG risks  
+- Explanation Agent – Generates human-readable explanations for ESG score changes  
+
+---
+
+## 🔧 Technical Architecture
+
+### Frontend
+- React.js  
+- Tailwind CSS  
+- Material UI  
+- Socket.IO  
+- Recharts / Chart.js  
+
+### Backend
+- Node.js  
+- Express.js  
+- Socket.IO  
+- JWT Authentication  
+
+### AI / ML
+- Python  
+- Scikit-Learn  
+- SpaCy  
+- BERT  
+- Multi-Agent System  
+
+### Automation & Orchestration
+- n8n  
+- Webhooks  
+- REST APIs  
+
+### Database, Cloud & Deployment
+- MongoDB Atlas  
+- Docker  
+- Vercel  
+- Render / AWS  
+- Hugging Face  
+
+---
+
+## 🔄 System Workflow
+
+1. Data ingestion from IoT sensors, APIs, ERP systems, CSVs, and databases  
+2. Data cleaning, normalization, and preprocessing  
+3. Parallel processing using multi-agent architecture  
+4. Predictive ESG risk scoring using historical and live data  
+5. Explainable AI generates insights behind ESG score changes  
+6. Live dashboard displays ESG scores, alerts, and compliance status  
+
+---
+
+## 📊 Feasibility & Viability
+
+- Real-time data ingestion from multiple sources  
+- Modular and scalable multi-agent architecture  
+- Machine-readable regulatory rules (SEBI BRSR, GRI, EU CSRD)  
+- Automated compliance checks reduce manual effort  
+- Easy extension for new metrics, regulations, and industries  
+
+---
+
+## 🌍 Impact & Benefits
+
+### Benefits
+- Reduced penalties and reputational damage  
+- Better strategic decision-making  
+- Scalable and future-ready ESG framework  
+
+### Impact
+- Proactive ESG risk management  
+- Continuous compliance monitoring  
+- Higher transparency and accountability  
+- Supports sustainability goals  
+
+---
+
+## 💼 Business Model
+
+### Target Customers
+Mid-sized enterprises in IT, manufacturing, retail, and energy sectors with mandatory ESG compliance requirements.
+
+### Revenue Model
+Subscription-based SaaS (monthly/yearly) with paid add-ons for customization and enterprise integrations.
+
+### Technology Advantage
+A unified ESG platform combining live data ingestion, AI-based risk detection, and explainable insights.
+
 ---
-title: ESG Monitoring System
-emoji: 📚
-colorFrom: indigo
-colorTo: gray
-sdk: docker
-pinned: false
----
 
-Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference
+## Deployment Notes
+
+Set these environment variables in your deployment platform before starting the API:
+
+```env
+MONGO_URI=mongodb+srv://<username>:<password>@<cluster-url>/?appName=<app-name>
+MONGO_DB_NAME=company_database
+MONGO_FIRM_ID=TECH001
+MONGO_COLLECTION=raw_firm_data
+```
+
+The backend already reads `config/.env` locally through `python-dotenv`, and cloud deployments should provide the same keys as platform environment variables. Change `MONGO_FIRM_ID` to switch the active company without passing query parameters.
+
+
+
+
+

config/.env.example

@@ -0,0 +1,3 @@
+MONGO_URI=
+MONGO_DB_NAME=
+MONGO_COLLECTION=

config/settings.py

@@ -0,0 +1,8 @@
+from pathlib import Path
+
+from dotenv import load_dotenv
+
+BASE_DIR = Path(__file__).resolve().parents[1]
+ENV_FILE = BASE_DIR / "config" / ".env"
+
+load_dotenv(ENV_FILE)

models/train.py

@@ -4,25 +4,12 @@ from sklearn.model_selection import train_test_split
 import joblib
 
 def train_model():
-    # Load dataset
-    df = pd.read_csv(
-        r"C:\Users\Mahek Bhatia\Desktop\ESG-Monitoring-System\outputs\agent4_final_output.csv"
-    )
+    df = pd.read_csv(r"C:\Users\Mahek Bhatia\Desktop\ESG-Monitoring-System\outputs\agent4_final_output.csv")
 
-    # Create label
     df["risk_label"] = (df["final_esg_risk_score"] >= 66).astype(int)
 
-    # 🔥 Use SAME features that will come from API CSV
-    features = [
-        "ESG_Score",
-        "E_Score",
-        "S_Score",
-        "G_Score",
-        "Board_Independence",
-        "Firm_Size"
-    ]
-
-    X = df[features]
+    # Drop non-feature columns
+    X = df.drop(columns=["risk_label"])
     y = df["risk_label"]
 
     X_train, X_test, y_train, y_test = train_test_split(

outputs/TECH001/TECH001_agent1_operational_output.csv

@@ -0,0 +1,11 @@
+Year,Industry_Type,E_Score,S_Score,G_Score,ROA,ROE,Net Profit Margin,Board_Independence
+2014,Technology,93.7,82.2,84.3,6.4,10.4,16.8,89.4
+2015,Technology,62.9,86.2,67.3,11.9,14.6,11.6,57.2
+2016,Technology,89.2,88.4,67.4,7.5,13.8,17.6,93.2
+2017,Technology,75.8,73.4,81.0,16.8,19.4,15.8,72.3
+2018,Technology,49.7,94.5,73.3,2.5,15.7,6.7,95.0
+2019,Technology,44.3,83.4,77.4,14.2,24.5,20.7,81.7
+2020,Technology,69.3,52.3,61.8,3.4,17.9,17.0,93.5
+2021,Technology,51.0,74.6,61.7,9.5,23.2,18.0,50.1
+2022,Technology,68.2,92.9,91.3,9.3,14.8,19.3,69.9
+2023,Technology,47.0,88.5,79.1,4.6,9.5,21.2,89.2

outputs/TECH001/TECH001_agent2_financial_output.csv

@@ -0,0 +1,11 @@
+Year,Industry_Type,E_Score,S_Score,G_Score,ROA,ROE,Net Profit Margin,Board_Independence,ESG_Score,financial_risk_score
+2014,Technology,93.7,82.2,84.3,6.4,10.4,16.8,89.4,86.73,0
+2015,Technology,62.9,86.2,67.3,11.9,14.6,11.6,57.2,72.13,0
+2016,Technology,89.2,88.4,67.4,7.5,13.8,17.6,93.2,81.67,0
+2017,Technology,75.8,73.4,81.0,16.8,19.4,15.8,72.3,76.73,0
+2018,Technology,49.7,94.5,73.3,2.5,15.7,6.7,95.0,72.5,33
+2019,Technology,44.3,83.4,77.4,14.2,24.5,20.7,81.7,68.37,33
+2020,Technology,69.3,52.3,61.8,3.4,17.9,17.0,93.5,61.13,33
+2021,Technology,51.0,74.6,61.7,9.5,23.2,18.0,50.1,62.43,33
+2022,Technology,68.2,92.9,91.3,9.3,14.8,19.3,69.9,84.13,0
+2023,Technology,47.0,88.5,79.1,4.6,9.5,21.2,89.2,71.53,33

outputs/TECH001/TECH001_agent3_compliance_output.csv

@@ -0,0 +1,11 @@
+Year,Industry_Type,E_Score,S_Score,G_Score,ROA,ROE,Net Profit Margin,Board_Independence,ESG_Score,financial_risk_score,E_Compliance,S_Compliance,G_Compliance,Overall_Compliance
+2014,Technology,93.7,82.2,84.3,6.4,10.4,16.8,89.4,86.73,0,Compliant,Compliant,Compliant,Compliant
+2015,Technology,62.9,86.2,67.3,11.9,14.6,11.6,57.2,72.13,0,Compliant,Compliant,Compliant,Compliant
+2016,Technology,89.2,88.4,67.4,7.5,13.8,17.6,93.2,81.67,0,Compliant,Compliant,Compliant,Compliant
+2017,Technology,75.8,73.4,81.0,16.8,19.4,15.8,72.3,76.73,0,Compliant,Compliant,Compliant,Compliant
+2018,Technology,49.7,94.5,73.3,2.5,15.7,6.7,95.0,72.5,33,Violation,Compliant,Compliant,Non-Compliant
+2019,Technology,44.3,83.4,77.4,14.2,24.5,20.7,81.7,68.37,33,Violation,Compliant,Compliant,Non-Compliant
+2020,Technology,69.3,52.3,61.8,3.4,17.9,17.0,93.5,61.13,33,Compliant,Violation,Compliant,Non-Compliant
+2021,Technology,51.0,74.6,61.7,9.5,23.2,18.0,50.1,62.43,33,Violation,Compliant,Compliant,Non-Compliant
+2022,Technology,68.2,92.9,91.3,9.3,14.8,19.3,69.9,84.13,0,Compliant,Compliant,Compliant,Compliant
+2023,Technology,47.0,88.5,79.1,4.6,9.5,21.2,89.2,71.53,33,Violation,Compliant,Compliant,Non-Compliant

outputs/TECH001/TECH001_agent4_final_output.csv

@@ -0,0 +1,11 @@
+Firm_ID,Year,Overall_Compliance,final_esg_risk_score,alert_level
+TECH001,2014,Compliant,0,Low
+TECH001,2015,Compliant,0,Low
+TECH001,2016,Compliant,0,Low
+TECH001,2017,Compliant,0,Low
+TECH001,2018,Non-Compliant,33,Warning
+TECH001,2019,Non-Compliant,33,Warning
+TECH001,2020,Non-Compliant,33,Warning
+TECH001,2021,Non-Compliant,33,Warning
+TECH001,2022,Compliant,0,Low
+TECH001,2023,Non-Compliant,33,Warning

outputs/TECH001_master.csv

@@ -0,0 +1,11 @@
+Year,Industry_Type,E_Score,S_Score,G_Score,ROA,ROE,Net Profit Margin,Board_Independence
+2014,Technology,93.7,82.2,84.3,6.4,10.4,16.8,89.4
+2015,Technology,62.9,86.2,67.3,11.9,14.6,11.6,57.2
+2016,Technology,89.2,88.4,67.4,7.5,13.8,17.6,93.2
+2017,Technology,75.8,73.4,81.0,16.8,19.4,15.8,72.3
+2018,Technology,49.7,94.5,73.3,2.5,15.7,6.7,95.0
+2019,Technology,44.3,83.4,77.4,14.2,24.5,20.7,81.7
+2020,Technology,69.3,52.3,61.8,3.4,17.9,17.0,93.5
+2021,Technology,51.0,74.6,61.7,9.5,23.2,18.0,50.1
+2022,Technology,68.2,92.9,91.3,9.3,14.8,19.3,69.9
+2023,Technology,47.0,88.5,79.1,4.6,9.5,21.2,89.2

requirements.txt

@@ -5,3 +5,8 @@ scikit-learn
 joblib
 python-multipart
 shap
+Flask
+requests
+python-dotenv
+pymongo
+python-socketio