The focus is the everyday information problems of architectural and BIM projects, building typologies such as offices, healthcare, and mixed-use, where room programs, specifications, and model quality all matter. The whole workshop follows one sample project that you progressively enrich, so every step builds toward a single finished result rather than disconnected exercises.
The workshop is hands-on: most of the time is spent building, not watching. We start by looking at a striking finished dashboard, the target, then spend the two days building the real, working version behind it. To keep everyone in step, each stage has a checkpoint file you can load to rejoin if a step breaks.
You work hands-on throughout: everyone cleans and merges data in Power Query and sets up their own pyRevit plugin. For the AI steps, the exact prompt used in Claude will be provided, and the reply it produced will be applied to the same prompt, with the result included together so the whole group stays in sync.
Checkpoint files, including the finished dashboard, mean no one is ever blocked, and anyone with a Claude account can run the prompts live as an extra.
Participants work individually, and the closing session is a shared discussion on applying the workflows to real projects.
Program
Day 1: Building and Structuring BIM Data
- Why project data matters and overview of the final health-score dashboard
- Power Query fundamentals: importing, cleaning, merging, appending, and building a refreshable workflow
- Workflow 1: Specification Database → Revit Keynotes
Clean and structure specification data in Power Query and export a Revit-ready keynote file - Workflow 2: Merging Revit Quantities with External Data
Combine Revit outputs with external datasets into a single validated dataset using repeatable Power Query steps
Day 2: Automation, Validation & Dashboard Integration
- Set up pyRevit and build a simple Revit plugin (hands-on)
Create a basic add-in button that links Revit to an external dashboard, with Claude-assisted coding - Model checking / QA-QC rules
Automate checks for missing fire ratings, marks, departments, and classification data - Space program validation
Compare model data against the project brief and measure area compliance - BIM Health Scorecard (AI-assisted workflow)
Follow a guided Claude prompt to generate a self-contained HTML dashboard from validated data - System integration
Connect the Revit plugin to the dashboard for one-click access from the model - Optional: live AI workflow
If available, participants can run Claude prompts themselves instead of following pre-generated outputs - Wrap-up
Full pipeline recap: from Revit and Power Query to AI-generated dashboard and plugin integration, and real-world applications
