💰 DeepSeek AI - Personal Financial Advisor & Expense Tracker

🚀 DeepSeek AI helps you track and analyze your personal finances by integrating AI-powered expense tracking, SQL generation, and financial insights. It’s a self-hosted AI-powered bookkeeper that securely manages your financial data without cloud dependency.

---

✨ Features

• 📂 CSV-Based Expense Tracking – Upload your bank transactions CSV to analyze spending.
• 📊 AI Categorization – DeepSeek AI intelligently classifies transactions (e.g., Groceries, Rent, Travel).
• 🧠 AI-Powered Insights – Get actionable financial advice based on your transactions.
• 🏦 Database Integration – Store transactions in PostgreSQL for historical analysis.
• 🔍 AI-Generated SQL Queries – Just ask natural language financial questions, and DeepSeek will generate & execute SQL queries to fetch insights.
• 🔄 Auto-Correcting AI Queries – If SQL fails, AI retries and refines the query until it works.
• 📈 Financial Visualization – Get bar charts, spending trends, and breakdowns automatically.
• 🔐 100% Local & Private – Runs on your own machine with Ollama & PostgreSQL (no cloud data sharing).

---

🛠️ Tech Stack

• AI Model: DeepSeek R1 (Running locally via Ollama)
• UI: Streamlit
• Database: PostgreSQL
• Backend: Python

---

🔧 Setup Instructions

1️⃣ Install Dependencies

pip install -r requirements.txt

2️⃣ Set Up PostgreSQL Database

Ensure you have PostgreSQL running locally and update your .env file with database credentials:

POSTGRES_HOST='your_db_host'
POSTGRES_PORT='5432'
POSTGRES_USER='your_db_user'
POSTGRES_PASSWORD='your_db_password'
POSTGRES_DB='your_db_name'

3️⃣ Run Ollama Locally

Make sure DeepSeek AI is available via Ollama:

ollama run deepseek-r1:8b

4️⃣ Start the Streamlit App

streamlit run app/main.py

---

🏗️ How It Works

1️⃣ Upload Your Transactions – Upload a CSV file of your bank transactions.
2️⃣ AI Categorization & Graphs – AI automatically categorizes transactions and shows insights with bar charts.
3️⃣ Store in Database – Click "Load Data to PostgreSQL" to store your transactions for historical analysis.
4️⃣ Ask AI Financial Questions – Example:

• "What was my biggest expense last month?"
• "How much did I spend on groceries this year?"
  5️⃣ AI Generates & Executes SQL – AI converts the question into SQL, runs it, and provides answers.
  6️⃣ Self-Correcting Queries – If the query fails, AI refines and retries until it works.

---

🎯 Example Queries

✅ "What was my biggest expense last month?" → AI generates:

SELECT MAX(amount), description FROM upload_transactions WHERE transaction_type = 'Expense' AND month_year = '2024-01';

✅ "How much have I spent on food?" → AI generates:

SELECT SUM(amount) FROM upload_transactions WHERE category = 'Food';

---

🔥 Future Improvements

• ✅ More AI-powered financial insights
• ✅ Interactive dashboard with filters
• ✅ Support for multiple banks & currencies
• ✅ Export customized financial reports

---

💡 Contribute & Feedback

I built this project to securely manage my finances using AI and help others do the same.
If you find it useful or have suggestions, feel free to contribute or open an issue!

📩 Let’s improve this together!

---

💙 If you found this useful, give it a ⭐ on GitHub and share your thoughts!