AI-Powered Real-Time Finance Tracker

🚀 Project Overview

Our AI-Powered Real-Time Finance Tracker is a cutting-edge financial management system that enables users to:

• Seamlessly upload transactions 📂
• Link bank accounts securely 🏦
• Gain AI-driven insights and personalized financial advice 🤖
• Visualize earnings, investments, and expenses 📊
• Future Finance Advisor for our benefits
• Interactive Pie Charts, Line Graphs, Bar Plots for our personal finance data

This project leverages Pathway Vector Store, Fetch.AI Agents (User Agent & Finance Advisor Agent), and integrates Next.js (frontend), TypeScript, and Flask (backend) to create a fully functional and intelligent financial tracking system.

---

🛠️ Tech Stack

Technology	Purpose
Next.js	Frontend framework for a responsive UI
TypeScript	Type-safe JavaScript for better maintainability
Flask	Lightweight backend for API interactions
Pathway Vector Store	Efficient transaction storage and retrieval
Fetch.AI Agents	Autonomous AI agents for financial recommendations

---

📈 Feature Flowchart

graph TD  
    A[User Uploads Transactions / Links Bank Account] -->|Data Stored| B[Pathway Vector Store]  
    B -->|Fetch.AI User Agent Processes Transactions| C[Fetch.AI Finance Advisor Agent]  
    C -->|Analyzes Data & Generates Insights| D[AI-Driven Financial Advice]  
    D -->|Displays Recommendations & Visualizations| E[Next.js Frontend Dashboard]  
    E -->|Triggers Email Alerts If Budget Exceeded| F[Real-Time Email Notifications]  

Loading

---

🔎 RAG (Retrieval-Augmented Generation) Pipeline

The RAG Pipeline ensures that financial insights are generated in real-time using both stored transaction history and external data sources. The following interactive flowchart visualizes the RAG process:

---

FrostHack2025

🚀 Running Instructions

✅ Method 1: Manually

1️⃣ Backend Setup (Flask & Fetch.AI Agents)

# Clone the repository  
git clone https://github.com/Aman071106/FrostHack2025.git  
cd FrostHack2025  

# Set up virtual environment  
python3 -m venv venv  
source venv/bin/activate  # On Windows: venv\Scripts\activate  

2️⃣ Environment Variables

• Create a .env file in the root directory and add the following:

GEMINI_API_KEY = AIzaSyD59g**************************52L2BAWKc  
DB_USER = ****  
DB_PASSWORD = ******** 
DB_HOST = ******
DB_PORT = **** 
DB_NAME = ****** 

3️⃣ Install Dependencies

pip install -r requirements.txt

4️⃣ Run Backend Agents

cd aibackend/agents
python rag.py
python user.py
cd ..
cd update_transactions
python csv_uploader.py

💻 Frontend Setup

✅ Option 1: Quick Testing (Streamlit App)

cd aibackend/app
streamlit run app.py

✅ Option 2: Full Setup (Next.js + MongoDB Auth)

# Setup Auth Backend
cd auth_backend
# Create a .env file with your MongoDB URI
# MONGODB_URI=mongodb+srv://<your-uri>

npm install dotenv mongodb express cors mongoose bcryptjs
node server.js
# Setup Frontend
cd ../../frontend
npm install next react react-dom

cd app
npm run dev

📦 Method 2: Via Docker

• 1. Streamlit app
• Create a .env file in the root directory and add the following:(or use -e tags)

GEMINI_API_KEY = AIzaSyD59g**************************52L2BAWKc  
DB_USER = ****  
DB_PASSWORD = ******** 
DB_HOST = ****** (ip address if machine localhost as docker localhost is diff or use `host.docker.internal`)
DB_PORT = **** 
DB_NAME = ****** 

  docker build -t st_app .

OR

  docker pull deadlyharbor/st_app

Then

 docker run -p 8080:8080 -p 8000:8000 -p 8501:8501 -p 5001:5001 --env-file .env st_app

---

📊 Dashboard Preview

🤖 Financify Agent

---

🎯 Future Enhancements

• 🏦 Multi-Bank Support (Integrate Plaid API for wider banking compatibility)
• 📡 Blockchain Integration (For secure financial transactions & tracking)

---

🤝 Contributors

• Aman Gupta - Backend & AI Development
• Harsh Yadav - Backend & AI Development
• Kunal Mittal - UI/UX, Frontend Development and Backend Development

---

⭐ Get Started Today and Plan Your Finance! 🚀

---

✅ What's New?

• Added RAG Pipeline Flowchart 📊
• Flowcharts are now interactive (GitHub renders MermaidJS)
• Updated Local Hosting Instructions for better clarity

Let me know if you want any more modifications! 🚀