An intelligent assistant that helps users query and analyze financial transactions using semantic similarity and vector embeddings.
It can understand natural language queries like “Show my top 5 expenses in September” and return relevant insights.
Build an AI-powered financial data assistant that can:
- 🧾 Generate synthetic financial transactions using AI
- 💾 Store them in a vector database using embeddings
- 💬 Answer natural language queries, such as:
“How much did I spend on food last month?” 
- 📊 Return the most relevant transactions or summarized insights
Financial-data/
│
├── .ipynb_checkpoints/
│   data/
│   ├── embeddings/
│   └── transactions.json
├── .env
├── .gitignore
├── Financial-Data-Assistant.ipynb
├── main.py
└── README.md             
🚀 How to Run the Project
git clone https://github.com/CodeWithDarshan04/Financial-Data-Assistant.git
cd Financial-Data-Assistant
uvicorn main:app --reload
    Category	                          Tools/Libraries
    Language	                          Python 
    Dummy Data Generator:                 Faker
    Embedding Model:	                  sentence-transformers
    API:	                              FastAPI
    Vector DB:	                          FAISS 
    Database:	                          JSON file (no real DB required)
How much did I spend on food last month?
Output: [ { "id": "txn_b2569544", "userId": "user_1", "date": "2025-06-09", "description": "UPI payment to Mitter-Kothari", "amount": 4947, "type": "Debit", "category": "Food", "balance": 141975, "similarity": 0.33, }
What’s my biggest expense in September
Output: { 'id': 'txn_8045a711', 'userId': 'user_2', 'date': '2025-03-15', 'description': 'UPI payment to Parmer-Deep', 'amount': 3138, 'type': 'Debit', 'category': 'Travel', 'balance': 272901, 'similarity': 0.308 } { 'id': 'txn_a5a046c5', 'userId': 'user_2', 'date': '2025-04-15', 'description': 'UPI payment to Rai-Keer', 'amount': 3033, 'type': 'Debit', 'category': 'Rent', 'balance': 26424, 'similarity': 0.304 }