🤖 The future of CAD (BIM) data processing in Construction is already here!

Instead of working with complex closed or parametric data (the quality of which we have to check in special BIM tools), in the future we work with open structured data, which is an ideal source for RAG, LLM, ChatGPT. To work with LLM models and ChatGPT you need properly prepared data.

Dataframe data obtained from various CAD (BIM) formats is the ideal fuel for modern tools. Pandas is ideal for LLMs due to its robust data processing, efficient indexing, and flexible formatting. It cleans, tokenizes, and normalizes data, supports advanced data retrieval, and structures data in LLM-compatible formats.

Examples of requests for using structured CAD (BIM) data in ChatGPT:

🤖Group the data in Dataframe by "Type Name" while summarizing 
the "Volume" parameter - you can show the result in any kind of graphs and documents

🤖 “Check the values in the volume parameter for all items in "Category"
 - "OST_Walls" and output the list of IDs with null values” - native IDs 
can be displayed later in Revit

🤖 “Check if "Category" "OST_Doors" has parameters that are responsible
for width and length and output element types that do not have such parameters”

⚡️ Instead of using a whole zoo of BIM tools, we now validate and process data directly in ChatGPT.

After getting the results in ChatGPT, we copy the resulting Python code into Pipeline in any Python IDE