https://github.com/chhitij/fin-advise-app
FinAdvise - AI-Powered Personal Finance Assistant
https://github.com/chhitij/fin-advise-app
genai-chatbot langchain langgraph multiagent
Last synced: about 1 month ago
JSON representation
FinAdvise - AI-Powered Personal Finance Assistant
- Host: GitHub
- URL: https://github.com/chhitij/fin-advise-app
- Owner: chhitij
- Created: 2025-11-16T19:11:21.000Z (8 months ago)
- Default Branch: main
- Last Pushed: 2025-11-16T20:27:06.000Z (8 months ago)
- Last Synced: 2025-11-16T21:18:18.335Z (8 months ago)
- Topics: genai-chatbot, langchain, langgraph, multiagent
- Language: Python
- Homepage: https://fin-advise-app.onrender.com/
- Size: 21.5 KB
- Stars: 0
- Watchers: 0
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: readme.md
Awesome Lists containing this project
README
💰 FinAdvise - AI-Powered Personal Finance Assistant
An intelligent financial advisor chatbot built with **LangGraph**, **LangChain**, and **OpenAI GPT** that helps users manage their finances, track expenses, analyze stocks, and receive personalized financial advice.
## LangSMith https://smith.langchain.com
## 🎯 Purpose
FinAdvise is designed to democratize financial literacy by providing:
- **Real-time stock price information** using Yahoo Finance
- **Expense tracking and budgeting** with automatic categorization
- **Personalized financial advice** based on user profiles
- **Context-aware conversations** that remember previous discussions
- **Human-in-the-loop safety** for high-risk financial decisions
- **LangSmith tracing** for monitoring and debugging AI interactions
## ✨ Key Features
### 1. **Stock Analysis**
- Get real-time stock prices for any US company
- Contextual follow-up questions (e.g., "Should I buy it?")
- Personalized investment advice based on risk tolerance and age
### 2. **Expense Tracking**
- Add expenses with natural language (e.g., "Add $50 for groceries")
- Automatic categorization (Housing, Food, Transportation, etc.)
- Track spending patterns over time
### 3. **Budget Management**
- View spending breakdown by category
- Compare income vs. expenses
- Get alerts when over budget
- See recent transaction history
### 4. **Financial Advice**
- Personalized recommendations based on your profile
- Risk-appropriate investment strategies
- Budget optimization tips
- Savings goal planning
### 5. **Smart Memory System**
- **Short-term memory**: Remembers context within conversation
- **Long-term memory**: Stores user profile and financial goals
- **Expense history**: Tracks all transactions
### 6. **Safety Features**
- **Human-in-the-Loop (HITL)**: Flags high-risk queries like "liquidate retirement"
- **Risk assessment**: Provides advice based on user's risk tolerance
- **Transparent tracking**: All AI interactions logged in LangSmith
## 🚀 Quick Start
### Prerequisites
- Python 3.8+
- OpenAI API key
- LangSmith API key (optional, for tracking)
### Installation
1. **Clone the repository:**
```bash
cd d:\genrative-ai-repo\stock-analysis
```
2. **Create conda environment:**
```bash
conda create -n finapp python=3.10
conda activate finapp
```
3. **Install dependencies:**
```bash
pip install -r requirements.txt
```
4. **Create `.env` file:**
```env
OPENAI_API_KEY=sk-your-openai-key-here
LANGSMITH_API_KEY=lsv2_pt_your-langsmith-key-here
LANGCHAIN_TRACING_V2=true
LANGCHAIN_PROJECT=finadvise-app
```
5. **Run the app:**
```bash
python app.py
```
6. **Open in browser:**
- Local: http://127.0.0.1:7860
- Public: Check terminal for Gradio share link
## 📊 Sample Questions
### Stock Analysis
```
✅ "What's Apple stock price?"
✅ "How much is TSLA trading at?"
✅ "Tell me about Microsoft stock"
✅ "Should I buy AAPL?" (after asking about Apple)
✅ "Is that a good price?" (follow-up question)
```
### Expense Tracking
```
✅ "Add $50 for groceries"
✅ "I spent $1200 on rent"
✅ "Spent $100 on utilities yesterday"
✅ "Add $45.99 for dinner"
✅ "I paid $80 for gas"
```
### Budget Summary
```
✅ "Show me a budget summary"
✅ "What's my spending breakdown?"
✅ "How much have I spent this month?"
✅ "Am I over budget?"
```
### Financial Advice
```
✅ "Should I invest in stocks or bonds?"
✅ "How much should I save each month?"
✅ "What's a good emergency fund amount?"
✅ "Is it a good time to invest?"
✅ "Should I pay off debt or invest?"
```
### User Profile
```
✅ "I'm 30 years old"
✅ "My income is $60,000 per year"
✅ "My goal is to save for retirement"
✅ "I have moderate risk tolerance"
✅ "I want to buy a house in 5 years"
```
### Follow-up Questions (Context Awareness)
```
✅ "What's Google stock?" → "Should I buy it?"
✅ "Add $200 for groceries" → "Show me my food expenses"
✅ "Tell me about Tesla" → "Is it worth investing in?"
```
### High-Risk Queries (HITL Triggers)
```
⚠️ "Should I liquidate my retirement account?"
⚠️ "I want to invest all my savings in crypto"
⚠️ "Should I sell everything?"
⚠️ "I want to use my entire portfolio for one stock"
```
## 🏗️ Architecture
### LangGraph Workflow
```
User Input
↓
[Intent Detection] ──→ Classifies: profile/stock/expense/budget/advice
↓
[Routing Logic] ──→ Selects appropriate node
↓
[Action Node] ──→ Executes task (fetch stock, track expense, etc.)
↓
[Response] ──→ Returns to user
```
### State Management
```python
FinanceState = {
"user_input": str, # Current message
"intent": str, # Detected intent
"data": dict, # Response data
"user_profile": dict, # Age, income, risk tolerance
"short_term_memory": dict, # Last stock, previous intent
"long_term_memory": dict, # Historical advice
"hitl_flag": bool, # Safety trigger
"expenses": list # Tracked expenses
}
```
## 🛠️ Technology Stack
| Component | Technology |
|-----------|-----------|
| **AI Framework** | LangChain, LangGraph |
| **LLM** | OpenAI GPT-4o-mini (fallback: gpt-5-nano) |
| **Stock Data** | yfinance |
| **UI** | Gradio (ChatInterface) |
| **Monitoring** | LangSmith |
| **Environment** | python-dotenv |
## 📈 LangSmith Tracking
View all AI interactions, prompts, and responses:
1. Go to https://smith.langchain.com
2. Select project: **finadvise-app**
3. View traces for:
- All LLM calls
- State transitions
- Token usage
- Latency metrics
## 🎨 Customization
### Change Theme
```python
# In app.py, line 437
dark_theme = gr.themes.Soft() # Options: Monochrome, Soft, Glass, Base
```
### Adjust LLM Temperature
```python
# In app.py, line 56
llm = ChatOpenAI(
temperature=0.3 # Lower = more consistent, Higher = more creative
)
```
### Add New Intent
1. Update `detect_intent()` prompt
2. Add new node function
3. Update routing in `get_next_node()`
4. Add conditional edge in graph builder
## 🔒 Security
- ✅ API keys stored in `.env` (never committed)
- ✅ High-risk queries flagged for human review
- ✅ User data stored in memory only (not persisted)
- ✅ All AI calls logged for audit trail
## 🐛 Troubleshooting
### "OPENAI_API_KEY not found"
- Check `.env` file exists in project root
- Verify key format: `OPENAI_API_KEY=sk-...`
### "Module not found" errors
```bash
pip install -r requirements.txt
```
### Stock data errors
- Yahoo Finance may rate-limit requests
- Use valid stock symbols (AAPL, TSLA, GOOGL, etc.)
### LangSmith not tracking
- Verify `LANGCHAIN_TRACING_V2=true` in `.env`
- Check API key is valid
- Ensure internet connection
## 📝 License
MIT License - feel free to use and modify!
## 🤝 Contributing
Contributions welcome! Areas for improvement:
- Add more financial indicators (P/E ratio, dividends)
- Persistent storage (database integration)
- Multi-language support
- Voice input/output
- Portfolio optimization algorithms
## 📧 Contact
For questions or support, open an issue on GitHub.
---
**Built with ❤️ using LangChain and LangGraph**