🛡️ Autonomous AI Compliance Platform

      **An agentic, LLM-powered compliance automation platform built for financial services organizations.** Automate regulatory gap analysis, policy auditing, real-time risk monitoring, and audit report generation — all orchestrated by a team of specialized AI agents.

---

📌 Overview

The Autonomous AI Compliance Platform replaces traditional, manual compliance workflows with an intelligent multi-agent system. It ingests regulatory documents (PCI-DSS, GDPR, etc.), compares them against internal policies using Retrieval-Augmented Generation (RAG), surfaces compliance gaps with risk scores, and generates industry-grade PDF audit reports — all with minimal human intervention.

Why This Platform?

Traditional Compliance	This Platform
Manual document review	AI-powered automated ingestion
Days to weeks per audit	Scan in minutes
Static, point-in-time reports	Real-time risk monitoring
Human error in gap analysis	LLM-driven reasoning with RAG context
Generic templates	Client-specific PDF audit packages

---

✨ Features

🤖 Multi-Agent Architecture (ReAct Pattern)

Four specialized AI agents collaborate using the Reason + Act paradigm:

Agent	Role	Tools Used
Regulatory Scout	Parses and indexes regulatory documents	Document Reader, RAG Indexer
Gap Analyst	Detects compliance gaps, scores risks	LLM Reasoning, Risk Scorer
Risk Sentinel	Monitors live transaction streams for violations	Rule-based Engine
Evidence Officer	Compiles and generates PDF audit reports	PDF Generator

🧠 AI & Knowledge Features

• RAG Pipeline: Documents are embedded and stored in ChromaDB for semantic retrieval during analysis
• Gemini 2.0 Flash: Powers all LLM reasoning — obligation extraction, gap detection, and auto-remediation suggestions
• Structured Extraction: LLM extracts regulatory obligations from raw text into structured schemas
• Auto-Remediation: Analyst agent optionally generates fix recommendations inline with each finding

📊 Dashboard & Reporting

• Real-time Compliance Score with trend tracking
• Risk Heatmap across regulatory domains
• Live Agent Activity Log with timestamped actions per client/document
• PCI-DSS 4.0 Checklist embedded in generated PDF reports
• One-click PDF download of full audit packages

🗂️ Document Management

• Multi-file upload — ingest PDFs and TXT files in batch
• Regulations page — manage your indexed knowledge base
• Reports page — list and download all generated audit reports

⚙️ Configurable Agent Settings

• Toggle auto-remediation per agent
• Adjust risk thresholds and monitoring parameters
• Live Gemini API status indicator
• All settings persisted via API

---

🏗️ Architecture

graph TB
    subgraph "Frontend Layer"
        UI[React Dashboard]
        FU[File Upload]
        RPT[Reports Viewer]
    end

    subgraph "API Gateway"
        API[FastAPI Server]
        MW[CORS Middleware]
    end

    subgraph "Agent Orchestration Layer"
        SCOUT[Regulatory Scout Agent]
        ANALYST[Gap Analyst Agent]
        SENTINEL[Risk Sentinel Agent]
        OFFICER[Evidence Officer Agent]
    end

    subgraph "Services Layer"
        LLM[LLM Service - Gemini 2.0 Flash]
        RAG[RAG Service - ChromaDB]
    end

    subgraph "Tools Layer"
        PDF[PDF Generator - FPDF2]
        READER[Document Reader - PyPDF]
        SCORER[Risk Scorer]
    end

    subgraph "Data Layer"
        CHROMA[(ChromaDB Vector Store)]
        FILES[(File System / Reports)]
    end

    UI --> API
    FU --> API
    RPT --> API
    API --> SCOUT & ANALYST & SENTINEL & OFFICER
    SCOUT --> LLM & RAG & READER
    ANALYST --> LLM & RAG & SCORER
    OFFICER --> PDF
    RAG --> CHROMA
    PDF --> FILES

Data Flow — Compliance Scan

Upload Documents → Scout Agent → ChromaDB (indexed)
                                      ↓
Trigger Scan → Analyst Agent → RAG Query → LLM Gap Detection → Risk Score
                                      ↓
           Evidence Officer → PDF Report Generation → Download

---

🗂️ Project Structure

autonomous-ai-compliance-platform/
│
├── client/                         # React 19 + Vite Frontend
│   ├── src/
│   │   ├── pages/
│   │   │   ├── Dashboard.jsx       # Main compliance dashboard
│   │   │   ├── Regulations.jsx     # Document upload & management
│   │   │   ├── Reports.jsx         # Download audit PDF reports
│   │   │   ├── RiskAnalysis.jsx    # Risk heatmap & analysis
│   │   │   ├── ActivityLog.jsx     # Agent activity timeline
│   │   │   ├── Settings.jsx        # Agent configuration
│   │   │   └── Login.jsx           # Authentication screen
│   │   ├── components/             # Reusable UI components
│   │   ├── lib/                    # API utilities & helpers
│   │   └── main.jsx
│   ├── Dockerfile
│   ├── nginx.conf
│   └── package.json
│
├── server/                         # FastAPI Python Backend
│   ├── agents/
│   │   ├── base.py                 # BaseAgent class (ReAct loop)
│   │   ├── scout.py                # Regulatory Scout Agent
│   │   ├── analyst.py              # Gap Analyst Agent
│   │   ├── sentinel.py             # Risk Sentinel Agent
│   │   └── evidence.py             # Evidence Officer Agent
│   ├── services/
│   │   ├── llm.py                  # Gemini LLM service wrapper
│   │   └── rag_service.py          # ChromaDB RAG service
│   ├── tools/                      # PDF generator, doc reader, etc.
│   ├── main.py                     # FastAPI app & all endpoints
│   ├── models.py                   # Pydantic data models
│   ├── database.py                 # SQLAlchemy setup
│   ├── requirements.txt
│   └── Dockerfile
│
├── docs/
│   └── architecture.md             # Full technical architecture doc
│
└── docker-compose.yml              # Full-stack deployment

---

🚀 Quick Start

Prerequisites

Requirement	Version
Python	3.10+
Node.js	18+
Gemini API Key	Get one free →
Docker (optional)	20+

---

Option A: Local Development

1. Clone the Repository

git clone https://github.com/your-username/autonomous-ai-compliance-platform.git
cd autonomous-ai-compliance-platform

2. Setup the Backend

cd server

# Create a virtual environment (recommended)
python -m venv venv
.\venv\Scripts\activate        # Windows
# source venv/bin/activate     # Linux / macOS

# Install Python dependencies
pip install -r requirements.txt

3. Configure Environment Variables

Create a .env file in the server/ directory:

GEMINI_API_KEY=your_gemini_api_key_here
DATABASE_URL=sqlite:///./compliance.db      # or PostgreSQL URL

Tip: Get your free Gemini API key at https://aistudio.google.com

4. Start the Backend Server

cd server
uvicorn main:app --reload

Backend runs at: http://localhost:8000
API docs (Swagger): http://localhost:8000/docs

5. Setup & Start the Frontend

Open a new terminal:

cd client
npm install
npm run dev

Frontend runs at: http://localhost:5173

---

Option B: Docker Compose (Recommended for Production)

Spin up the entire stack — backend, frontend, PostgreSQL, ChromaDB, Redis, and n8n — with a single command:

# Set your Gemini API key
export GEMINI_API_KEY=your_key_here   # Linux/macOS
$env:GEMINI_API_KEY="your_key_here"   # Windows PowerShell

# Start all services
docker-compose up --build

Service	URL
Frontend (React)	http://localhost:3000
Backend (FastAPI)	http://localhost:8000
ChromaDB	http://localhost:8001
n8n Automation	http://localhost:5678
PostgreSQL	localhost:5432
Redis	localhost:6379

---

🖥️ Using the Platform

1. Login — Open the app and sign in (demo accepts any credentials)
2. Check Status → Go to Settings to verify Gemini API is connected
3. Upload Documents → Go to Regulations → Upload one or more PDF/TXT policy files
4. Run a Compliance Scan → Click Trigger Scan on the Dashboard
5. Monitor Progress → Watch the Agent Activity Log update in real time
6. Review Findings → Explore the Risk Analysis page for gap details
7. Download Report → Go to Reports → Download your PDF audit package

---

📡 API Reference

Endpoint	Method	Agent	Description
/api/dashboard	GET	—	Dashboard metrics & compliance score
/api/agents/ingest	POST	Scout	Upload & index regulatory documents
/api/agents/scan	POST	Scout + Analyst	Trigger full compliance scan
/api/agents/analyze	POST	Analyst	Analyze a specific policy text
/api/agents/monitor	POST	Sentinel	Check transactions for violations
/api/agents/report	POST	Officer	Generate a PDF audit report
/api/agents/activity	GET	All	Fetch aggregated agent activity logs
/api/reports	GET	—	List all generated reports
/api/reports/download/{filename}	GET	—	Download a specific PDF report
/api/status	GET	—	Check Gemini API connectivity
/api/settings	GET/POST	—	Read/update agent configuration

Full interactive API docs available at http://localhost:8000/docs (Swagger UI).

---

🧩 Technology Stack

Layer	Technology	Purpose
Frontend	React 19 + Vite	Dashboard UI
Styling	TailwindCSS 3 + Framer Motion	Glassmorphism UI & animations
Charts	Recharts	Risk and compliance charts
Backend	FastAPI + Uvicorn	REST API & agent orchestration
LLM	Google Gemini 2.0 Flash	AI reasoning & text generation
Vector DB	ChromaDB	RAG semantic search
Relational DB	SQLite (dev) / PostgreSQL (prod)	Compliance data persistence
PDF Generation	FPDF2	Audit report creation
PDF Parsing	PyPDF	Document ingestion
Cache	Redis	API response caching
Workflow	n8n	Automation pipelines
Container	Docker + Docker Compose	Deployment

---

📋 Generated PDF Report Structure

Each audit report contains:

1. Cover Page — Client name, assessment date, compliance score, posture rating
2. Executive Summary — Requirements assessed, requirements met, gaps identified, critical issues
3. PCI-DSS 4.0 Checklist — All 14 requirements with PASS/FAIL status
4. Detailed Findings — Per-finding severity, description, and remediation steps
5. Recommendations — Prioritized action items with suggested timelines

---

🛠️ Troubleshooting

Gemini API not connected?

• Verify GEMINI_API_KEY is set in server/.env
• Restart the server: uvicorn main:app --reload
• Check the Settings page for live status

ChromaDB errors on first run?

• ChromaDB will auto-initialize its local DB on first document upload
• Ensure the server/chroma_db/ directory is writable

PDF not downloading?

• Verify the server/reports/ directory exists and has write permissions
• Check the FastAPI logs for PDFGenerator errors

Multi-file upload not working?

• Ensure python-multipart is installed (pip install python-multipart)
• Confirm you're on the latest frontend code and the server has been restarted

---

🤝 Contributing

Contributions are welcome! Please:

1. Fork the repository
2. Create a feature branch (git checkout -b feature/your-feature)
3. Commit your changes (git commit -m 'Add: your feature description')
4. Push to the branch (git push origin feature/your-feature)
5. Open a Pull Request

---

📄 License

This project is licensed under the MIT License — see the LICENSE file for details.

---

👤 Author

Built with ❤️ for the financial compliance industry.
Powered by Google Gemini · FastAPI · ChromaDB · React

Built With

• css
• dockerfile
• html
• javascript
• python