📊 Virtual Accounting Chatbot with FastMCP

An AI-powered Virtual Accounting Assistant built with FastMCP, local LLMs, and Streamlit.

This project demonstrates how the Model Context Protocol (MCP) can be used to build an intelligent accounting chatbot that understands natural language queries, retrieves financial data, and generates official accounting documents using local language models.

---

🔍 Table of Contents

• About the Project
• Features
• System Architecture
• Installation
• Running the Application
• UI Preview
• Project Structure
• Example Use Cases
• Technologies Used
• License

---

📌 About the Project

The Virtual Accounting Chatbot with FastMCP is an AI-based chatbot designed to assist with accounting and financial operations through natural language interaction.

It enables users to:

• Ask accounting-related questions
• Retrieve financial or taxpayer information
• Generate official documents such as petitions or reports
• Interact with structured tools via MCP instead of plain text responses

The project uses FastMCP to connect LLM reasoning with custom accounting tools, and Streamlit to provide a simple and interactive user interface.

---

🚀 Features

• 🤖 AI-powered accounting chatbot
• 🔗 Model Context Protocol (MCP) integration
• 🧠 Local LLM support (e.g., Ollama)
• 🖥️ Streamlit-based user interface
• 📄 Automatic document and petition generation
• 📊 Financial data querying via natural language
• 🛠️ Modular and extensible tool-based architecture

---

🧠 System Architecture

The system consists of four main components:

1. FastMCP Server

   • Acts as a bridge between the LLM and accounting tools
   • Exposes structured tools instead of raw text prompts

2. Local LLM (e.g., Ollama)

   • Understands user queries
   • Decides which MCP tools to call

3. Streamlit UI

   • Provides a chat-based interface
   • Displays responses and generated documents

4. Database (Optional)

   • Stores accounting data, query history, or mock financial records

---

🛠️ Installation

1️⃣ Clone the Repository

git clone https://github.com/iremclk812/Virtual-Accounting-Chatbot-with-FastMCP.git
cd Virtual-Accounting-Chatbot-with-FastMCP

2️⃣ Create a Virtual Environment

python -m venv venv
source venv/bin/activate      # macOS / Linux
venv\Scripts\activate         # Windows

3️⃣ Install Dependencies

pip install -r requirements.txt

---

▶️ Running the Application

Start the FastMCP Server

python mcp_server.py

Launch the Streamlit Interface

streamlit run chatbot_gui.py

Then open your browser at:

👉 http://localhost:8501

---

🖼️ UI Preview (Conceptual)

Below is a conceptual representation of the chatbot interface:

┌───────────────────────────────────────────┐
│ Virtual Accounting Chatbot                │
├───────────────────────────────────────────┤
│ User: Show my annual financial report     │
│ AI:   Please specify the year.            │
│                                           │
│ User: 2025                                │
│ AI:   Here is your 2025 financial summary │
│                                           │
├───────────────────────────────────────────┤
│ Tip: You can say "generate a petition"    │
└───────────────────────────────────────────┘

You can add real screenshots by placing images in the repository and linking them here.

---

📁 Project Structure

.
├── client/
│   └── assets/                # UI images or static files
├── db/
│   └── data.db                # Database (optional)
├── tools/
│   └── accounting_tools.py    # MCP accounting tools
├── mcp_server.py              # FastMCP server
├── chatbot_gui.py             # Streamlit UI
├── requirements.txt
├── README.md

---

💡 Example Use Cases

• “Show my tax obligations for this year.”
• “Get my current bank account balance.”
• “Summarize expenses by category.”
• “Generate an official accounting petition.”
• “Prepare a financial report for Q1.”

---

🧰 Technologies Used

• Python
• FastMCP (Model Context Protocol)
• Streamlit
• Local LLMs (e.g., Ollama)
• SQLite / JSON (optional storage)

---

📄 License

This project is licensed under the MIT License. See the LICENSE file for more details.

---