This project is about developing a locally hosted Large Language Model (LLM) based intelligent assistant using Python and the Model Context Protocol (MCP). The main aim of this project is to allow users to interact with an ai assistant without depending on cloud services, which helps in improving privacy And reducing cost and having better control over data.

The system is developed using Python and integrates a local LLM through Ollama. The asistant can process user queries, generate context aware responses and perform different tasks using MCP based communication. the MCP server acts like a bridge between the language model And external tools, allowing the assistant to access additional functionalities in a structured way.

The project was built using Kiro, which helped a lot in coding debugging and managing the projectWith the help of Kiro, the development process became much faster and easier, making it possible to quickly integrate different components.

one of the important features of this project is that it runs completely on a local machine, so it does not require continous internet connection. This makes it useful for students, researchers and organizations that need secure and private AI solutions. The architecture is modular, so new tools and custom features can be added easily through the MCP framework.

Overall, this project demonstrates the practical use of local AI models, protocol based tool integration and intelligent automation. It also shows how open source ai technologies can be used to create cost effective and privacy focused solutions for real world applications.