Financial Assistant Application

A comprehensive financial advisory system that provides personalized financial advice, investment opportunities, and account management based on user profiles and transaction history. Dataset: https://www.kaggle.com/datasets/priyamchoksi/credit-card-transactions-dataset

Table of Contents

1. Overview
2. Features
3. Architecture
4. Components
5. Setup
6. Usage
7. Security
8. Data Storage
9. API Integration

Overview

The Financial Assistant is a Streamlit-based web application that combines AI-powered financial advice with user profile management and transaction analysis. It uses LangChain, Ollama, and Tavily Search to provide personalized financial recommendations.

Features

User Management

• Secure login and signup system
• User profile management
• Session-based authentication
• Profile persistence using YAML configuration

Financial Analysis

• Transaction history analysis
• Account summary generation
• Spending pattern analysis
• Category-based expense tracking

Investment Research

• Location-specific investment opportunities
• Risk tolerance-based recommendations
• Market trend analysis
• Tax implications and benefits

AI-Powered Advice

• Personalized financial recommendations
• Context-aware responses
• Conversation history tracking
• Real-time streaming responses

Architecture

Core Components

1. User Interface (Streamlit)

   • Interactive web interface
   • Real-time chat interface
   • Profile management forms
   • Document upload system

2. AI Engine

   • LangChain for agent management
   • Ollama for LLM processing
   • Tavily Search for investment research
   • Custom callback handlers

3. Data Management

   • YAML-based configuration
   • Transaction data processing
   • Investment records storage
   • Conversation history tracking

Data Flow

1. User input → Streamlit interface
2. Query processing → LangChain agent
3. Tool execution → External APIs
4. Response generation → LLM
5. Result storage → Local files
6. Response display → User interface

Components

Main Application (main.py)

• Streamlit application setup
• User authentication
• Profile management
• Financial advice generation
• Investment opportunity search
• Conversation handling

Server Component (mcp_server.py)

• FastAPI server implementation
• API key validation
• Query processing
• Context-aware responses

Configuration (config.yaml)

• User credentials
• Profile information
• System settings

Setup

Prerequisites

• Python 3.8+
• Ollama running locally
• Tavily API key
• Required Python packages (see requirements.txt)

Installation

1. Clone the repository
2. Install dependencies:

   pip install -r requirements.txt
   

3. Create .env file with API keys:

   TAVILY_API_KEY=your_api_key
   

4. Start Ollama service
5. Run the application:

   streamlit run main.py
   

6. setup using uv and run:

   uv run python -m streamlit run main.py
   

Usage

User Authentication

1. Sign up with profile information
2. Log in with credentials
3. Update profile as needed

Financial Advice

1. Enter queries in the chat interface
2. Upload financial documents
3. View personalized recommendations
4. Track conversation history

Investment Research

1. Search for opportunities
2. View location-specific results
3. Analyze market trends
4. Get personalized recommendations

Security

Authentication

• Password hashing using SHA-256
• Session-based authentication
• API key validation
• Secure credential storage

Data Protection

• Environment variable management
• Secure file handling
• Input validation
• Error handling

Data Storage

User Data

• Stored in config.yaml
• Encrypted passwords
• Profile information
• Session data

Investment Records

• Stored in investment_records/
• JSON format
• Timestamp-based naming
• Complete context storage

Transaction Data

• CSV file processing
• In-memory caching
• Real-time analysis
• Historical tracking
• create a folder data/###.csv (Download from https://www.kaggle.com/datasets/priyamchoksi/credit-card-transactions-dataset)

API Integration

Tavily Search

• Investment research
• Market analysis
• Location-specific queries
• Real-time data retrieval

Ollama

• LLM processing
• Response generation
• Context management
• Streaming support

FastAPI Server

• Query processing
• Context handling
• Response formatting
• API key validation