FDX Smart Consent Manager 🚀

The FDX Smart Consent Manager is a full-stack AI-powered solution that demonstrates how financial data sharing can be made safer, more transparent, and user-friendly. It leverages Model Context Protocol (MCP) and Azure OpenAI to create a smart assistant that dynamically analyzes FDX consent APIs and explains them in simple terms.

---

🔧 Tech Stack

Layer	Technology
Frontend	React + TypeScript + Tailwind CSS
Backend	Spring Boot (Java 17)
AI Engine	Azure OpenAI (GPT-4o, GPT-4-turbo)
Protocol	MCP - Model Context Protocol
APIs	FDX v5.3 (Consent APIs)

---

✨ Features

• 🔍 Smart Consent Assistant: Conversational UI that simplifies consent language for users and developers.
• 🤖 Model Context Protocol: Dynamically builds a context chain using API specs, examples, and responses to ground LLMs.
• 🔐 FDX API Integration: Real-time request/response tracing for financial data consent.
• ⚡ Azure OpenAI Integration: Uses GPT-4o via Azure with deployment, key, and endpoint-based access.

---

🧠 Architecture Overview

!image

---

📁 Project Structure

fdx-smart-consent-manager/
├── backend/               # Spring Boot backend
│   └── src/main/java/com/crm/backend/mcp
│       ├── McpController.java
│       ├── McpService.java
│       ├── ModelRunner.java
│       └── FdxAdapter.java
├── ui/                    # React frontend
│   ├── pages/chat-assistant.tsx
│   ├── lib/utils.ts
│   └── public/
└── docs/
    └── architecture.png   # System architecture image

---

🚀 Getting Started

1. Clone the repo

git clone https://github.com/fdxhackathonpalteam/fdx-smart-consent-manager.git
cd fdx-smart-consent-manager

2. Backend Setup

cd backend
# Set environment variables or application.yml
./mvnw spring-boot:run

Make sure application.yml includes:

spring:
  ai:
    openai:
      endpoint: https://your-azure-openai-endpoint/
      api-key: YOUR_API_KEY
      deployment-id: gpt-4o
      api-version: 2024-12-01-preview

3. Frontend Setup

cd ui
npm install --legacy-peer-deps
npm run dev

Visit: http://localhost:3000

---

💡 MCP: Model Context Protocol

This project implements the MCP standard to:

• Collect and format API context
• Construct structured prompts
• Feed these prompts into Azure OpenAI Chat Completion API
• Return natural-language summaries or clarifications

---

📬 Sample Request

curl -X POST http://localhost:8080/api/mcp/ask   -H "Content-Type: application/json"   -d '{"question": "Explain what this consent means", "apiName": "GetConsent", "apiPayload": "{...}"}'

---

🤝 Contributors

• Suman Devarasetti
• NarasimhaMurthy Kapa
• Arvind Badri
• Mahesh Kotachery
• Vidya Parameswaran
• Venkata Kanumuri
• Parthiban Sivakumar

---

🧠 Inspiration

Built for the FDX Hackathon 2025, this project showcases how emerging AI standards like MCP + RAG can improve trust and transparency in Open Finance ecosystems.

---

📜 License

MIT License — use it, build on it, make it better!