You’re in a BIM coordination meeting, discussing a client submission deadline, unsure the team will make it. Ten to fifteen people are doing manual Autodesk Revit work that two or three people could clear in a day or two with the right tool. Everyone on the team asks you if there’s any way to fix this with automation.
Then, at the worst possible time, a Revit model update forces a full sheet resubmission, and the team’s still at it the next morning.
That’s the moment it hits you — automation would have stopped that night from happening.
You want to change the situation — same as I did — but nothing off the shelf quite fits how your Revit project runs. So here’s the solution I’m building right now.
1. Where the Hours Actually Go
Modelling complex parametric families, managing lakhs of element data properties, taking care of model health, and exporting — all these tasks have some standard Revit solutions and some third-party plugins. Dynamo and generic custom plugins can help, but what’s often missing is a script or plugin built for your specific task and workflow. For that, you have no real hope of automation — it’s manual, repetitive work instead.
2. The Tools I Tried Before
Dynamo and Dynamo Player worked, but crashed — with complex families and lakhs of data, a crash risked the whole model.
Rolling it out to the team meant training everyone, and some hit the same crashes a few times a week. Then I tried a Third Party trial plugin — it worked perfectly, but budget concerns made it hard to continue the subscription. So I googled some Python code and tried implementing it inside Dynamo.
With Claude’s help, I found out that going from Revit to Dynamo and back to Revit was a long, complex route that took more time — and was the main reason for the model crashes.
It suggested using Python code directly in the PyRevit shell instead. Surprisingly, that reduced the crashes to nearly zero.
3. When I Stopped Searching and Started Building
I found a problem I’d only half-solved — so I started building my own plugin…
This was all before AI tools like ChatGPT and Claude came along. I kept thinking: why can’t I build my own plugins for my own specific, native use?
PyRevit’s Python shell already existed; what I lacked was time to also become a Python developer on top of knowing the model and the drawings. Claude closed that gap — I bring the concept, Claude brings the code, tested together on a dummy project through several rounds of errors until it holds. Deployment became one file drop to the shared server location — the team gets the new tool on their ribbon after a simple reload, three clicks.
As I mentioned above, I have built my own PyRevit custom extension called BIM Minds. It has almost eliminated half of my day-to-day workload and made my work much easier.
4. What Changed on My Desk
It hasn’t flipped my work upside down, but a task that used to be a real burden closes in a day or two. No developer needed. It sits deliberately between a Dynamo script and a full commercial plugin. The real shift is being able to build a tool per person, per task, because I understand each person’s specific pain point — and a tool that proves itself travels to other projects. I don’t carry the old worry anymore.
Summary
This is the story of one of the biggest changes in my day-to-day as a BIM coordinator — going from chasing outdated Dynamo scripts and crash-prone plugins to building my own PyRevit extension, BIM Minds, with Claude’s help.
It didn’t take a developer, a big budget, or months of learning to code. Just the right approach, and a tool built for the exact problem I understood best.
Next up: how to actually set up and build a PyRevit extension, step by step.
Leave a Reply