This project demonstrates the ability to communicate with IFC files using natural language.
Powered by a large language model (LLM), the agent enables users to query and modify their IFC files.
For example, you can ask questions like:
Or ask to modify files:
This is an ongoing chat. You can also ask how your data was received, modified, or request further details.
The application leverages the Thatopen engine and a template app to provide a basic IFC file viewer that is extended with an AI Agent.
The application leverages the langgraph, azure and ifcopenshell to compose backend/the agent itself.
Any queries you make to the chatbot are saved anonymously, without any connection to your identity or account.
This is done to investigate which types of queries are being fulfilled and to improve the system.
Uploaded IFC files are not stored: as soon as your application session ends they are completely deleted from the server.
Please use one of the secure providers below to log in and try out the app:
Login with Google Login with Microsoft