INTELICREDIT: Next-Gen AI Corporate Credit Appraisal Engine

InteliCredit is a state-of-the-art enterprise lending platform designed for the Indian corporate market. It combines Deep Explainable AI (XAI), Graph Neural Networks (GNN), and Databricks Unity Catalog integration to automate the preparation of high-fidelity Credit Appraisal Memos (CAM).

• Neural Logic: Transparent ML decisioning using Gradient Boosting Regressors.
• Feature Attribution: Bar chart visualizations showing the contribution of each financial cluster (Liquidity, Leverage, etc.) to the Probability of Default (PD).
• Automated Justification: Natural language risk drivers generated per-entity.

• Network Analysis: Uses Graph Neural Networks to detect circular trading patterns and promoter money loops.
• Shell Company Guard: Identifies suspicious transaction clusters and round-tripping activities.

• GST vs Bank Analysis: Heuristics to detect "Window Dressing" by cross-validating GST filings against actual bank flows.
• GST-3B vs 2A Mismatch: Built-in logic for GST governance checks.

• Enterprise Data Sovereignty: Native support for three-level namespaces (catalog.schema.table).
• Unified Governance: Secure ingestion from governed SQL warehouses and data lakes.

• Backend: FastAPI, Python 3.12, PyPDF, WeasyPrint, Scikit-Learn.
• Frontend: Next.js 16 (Turbopack), Tailwind CSS, Recharts, Framer Motion.
• Identity: Firebase Google Auth (Stateless ID Token Verification).
• Intelligence: Google Gemini API (LLM), DuckDuckGo (Research).

/InteliCredit
|-- ai/                          # AI/ML Core (XAI, GNN, Gemini)
|-- api/                         # FastAPI Services & Pipeline (Stateless)
|-- cam_generator/               # Jinja2 PDF Templates & Logic
|-- document_parser/             # PDF/CSV Extraction
|-- financial_analyzer/          # Indian Context Heuristics (GST & Bank)
|-- frontend-nextjs/             # Next.js Dashboard
|-- risk_engine/                 # 5Cs Scoring Logic
|-- storage/                     # In-memory session data & CAM Output

pip install -r requirements.txt
cp .env.example .env
# Set GEMINI_API_KEY and FIREBASE_SERVICE_ACCOUNT_JSON
uvicorn api.main:app --reload

cd frontend-nextjs
npm install
npm run dev

Run the integrated verification suite:

python verify_phase6.py

This validates the end-to-end pipeline from Databricks ingestion to XAI report generation.

MIT License - Developed for the Intelli-Credit Challenge 2026.