CHAPTERS 1-12

⚡️ Free n8n CAD (BIM) workflow:
AUTOMATIC COST ESTIMATION FOR GROUPS AND THE ENTIRE PROJECT WITH LLM (works with Revit, IFC, DWG)

This is a ready-made solution on n8n that allows you to automatically calculate the cost of different groups of elements and the entire construction project based on data from Revit, IFC, or DWG files. It uses any LLM (such as Grok, OpenAI, Anthropic) — they automatically classify materials, search for prices in databases and online sources, and generate reports in a convenient format.

How does it work?
1️⃣ Revit/IFC/DWG are automatically converted into a structured DataFrame
2️⃣ LLM determines the grouping parameters, classifies materials, and finds current prices in databases and online
3️⃣ Reports and dashboards are generated automatically — visualizations and tables appear without manual routine work

The input is an RVT or IFC project, and the output is a finished document with a detailed analysis of groups by cost. At the same time, any parameter in the open workflow can be customized for your project (grouping process, country for prices, LLM API keys, or your own models). The average processing speed for one group of elements is 5 to 15 seconds, depending on the LLM you choose.

What does this mean for the market?
Traditional roles—BIM coordinator, BIM manager, estimator, designer—are gradually ceasing to be “manual data operators.” The key competency is no longer how to manually check a table and find the right group, but how to correctly compile a prompt, adjust process parameters, and analyze the result.

Real-life example:
Using ChatGPT, it took about 30 minutes to estimate the cost of a standard Revit project from Autodesk (rac_basic_sample.rvt). The total cost of the rac_basic_sample project is €647,700, with glass and walls accounting for the lion's share. The workflow even calculated the cost of the “vase on the table,” estimating it at €70. The XSLS and the HTML report received after the workflow is completed can be found in our telegram group  https://t.me/datadrivenconstruction/2941).

📥 Link to the workflow:
🔗 GitHub: https://github.com/datadrivenconstruction/cad2data-Revit-IFC-DWG-DGN-pipeline-with-conversion-validation-qto
Workflow file: n8n_Construction_Price_Estimation_with_LLM.json

Instructions inside the repo — import the workflow into n8n ➡️ specify the file, grouping, and country ➡️ add the LLM API keys ➡️ and run. Everything works out of the box, but the accuracy of the prompts can be easily adjusted to suit your experience. Advice: the highest recognition and classification accuracy in my tests was given by Claude 4 Opus and Grok4.

P.S. If anyone has RVT or IFC projects with already calculated costs, please write to me. I will be happy to test them on real data and fine-tune the prompts for your case.

♻️ Feel free to share this post with colleagues who are struggling with data quality and calculations, perhaps the headache and routine of the task can be solved by LLM