The Manual Tax: What AEC Loses Without CI/CD

The State of AEC Automation

Walk into any software company today, and you’ll find automated testing, continuous integration🔁CI/CDAutomated build, test, and deployment pipelines.View in glossary, and deployment pipelines that move code from a developer’s laptop to production with minimal human intervention. Now walk into an AEC firm’s BIM🏛️BIMIntelligent 3D model-based design for buildings.View in glossary department.

What do you see? Someone manually uploading Revit🏠RevitAutodesk's BIM software for architecture and construction.View in glossary files to BIM 360🔵BIM 360Legacy Autodesk construction platform (predecessor to ACC).View in glossary. A coordinator clicking through the Autodesk Platform Services☁️APSAutodesk Platform Services - cloud APIs for CAD/BIM automation.View in glossary web interface to check translation⚙️Translation JobBackground process converting CAD files to viewable formats.View in glossary status. An email chain deciding whether a model is “ready” for the next phase.

This is the manual tax — the cumulative cost of human intervention in processes that could be automated.

What Manual Processes Actually Cost

Time

Consider a typical model coordination workflow📈WorkflowAutomated process triggered by events.View in glossary:

1. Designer saves model locally (1 min)
2. Opens browser, navigates to ACC🏗️ACCAutodesk's construction management platform.View in glossary (1 min)
3. Uploads file, waits for transfer (5-15 min depending on size)
4. Triggers translation (1 min + wait)
5. Checks status periodically (10-30 min of interrupted work)
6. Emails team that model is ready (2 min)
7. Each team member downloads and opens (5 min × N people)

That’s 30-60 minutes per upload, multiplied by dozens of uploads per week, across multiple team members. The time adds up to weeks of lost productivity per year.

Errors

Manual processes introduce errors at every step🔧STEPISO standard for 3D CAD data exchange.View in glossary:

• Wrong file version uploaded
• Translation started with wrong parameters
• Notification sent before translation completed
• Team member missed the email and used outdated model

In software development, these would be caught by automated tests. In AEC, they’re often discovered days or weeks later during coordination meetings.

Context Switching

The most insidious cost is cognitive. When a designer has to stop their creative work to:

1. Remember where to upload
2. Navigate through web interfaces
3. Check back periodically for completion
4. Compose and send notifications

…they lose the mental context of their design work. Studies suggest it takes 23 minutes to fully regain focus after an interruption.

What Software Development Learned

The software industry went through this exact evolution. In the 1990s, “builds” were manual processes — someone would compile the code, run some tests by hand, and decide if it was ready to ship.

Then came continuous integration. The rules were simple:

1. Automate everything repeatable — if a human does it the same way twice, a computer should do it
2. Fail fast — catch errors in minutes, not days
3. Single source of truth — one process, documented in code, that everyone follows
4. Observability — know the status of everything at any moment

The results were transformative. Teams that adopted CI/CD shipped faster, with fewer bugs, and with happier developers.

Bringing CI/CD to APS

This is why I built RAPS🌼RAPSRust CLI for Autodesk Platform Services.View in glossary. The Autodesk Platform Services APIs are powerful and comprehensive — they expose everything needed to automate AEC workflows. But they’ve been accessible primarily to developers comfortable with REST🌐RESTWeb service architecture style using HTTP.View in glossary APIs and OAuth🔐OAuthIndustry-standard authorization protocol used by APS.View in glossary flows.

RAPS bridges that gap with a command-line💻CLIText-based interface for running commands.View in glossary interface that makes APS automation🤖AutomationReplacing manual processes with software.View in glossary accessible:

# This is what a CI/CD pipeline looks like
raps auth test
raps object upload my-bucket model.rvt
raps translate start $URN --format svf2 --wait
raps translate manifest $URN --output json

Four commands. No clicking. No waiting. No email chains.

Real-World Automation Example

Here’s what a complete automated workflow looks like in practice:

# GitHub Actions workflow
name: Model Processing Pipeline
on:
  push:
    paths: ['models/**/*.rvt']

jobs:
  process:
    runs-on: ubuntu-latest
    steps:
    - uses: actions/checkout@v4
    
    - name: Setup RAPS
      run: curl -fsSL https://rapscli.xyz/install.sh | sh
    
    - name: Process Changed Models
      run: |
        # Find changed models
        for model in $(git diff --name-only HEAD~1 HEAD | grep '\.rvt$'); do
          echo "Processing $model..."
          
          # Upload to cloud
          raps object upload project-models "$model"
          
          # Start translation  
          URN=$(raps object urn project-models "$model")
          raps translate start $URN --format svf2 --wait --timeout 1800
          
          # Generate thumbnail
          raps translate thumbnail $URN --width 400 --height 400
          
          # Notify team
          slack-notify "✅ $model processed and ready for review"
        done

The ROI of Automation

When this runs in a GitHub Actions🐙GitHub ActionsGitHub's built-in CI/CD platform.View in glossary workflow, it happens automatically every time a model changes. The team is notified through Slack or Teams. Errors are caught immediately. The status is visible in a dashboard, not someone’s inbox.

The Zero-Click Deployment Reality

With proper automation, your “time to deploy” goes from 20 minutes of clicking to 0 minutes. You push code, you go get coffee, and your models are processed and distributed.

No more:

• Forgetting to upload the latest version
• Uploading debug builds to production
• Manual coordination of file availability
• Context switching between design and deployment tasks

The Path Forward

Adopting CI/CD for APS workflows doesn’t require rewriting your entire process overnight. Start small:

1. Automate one workflow — pick the most painful manual process
2. Measure the improvement — time saved, errors avoided
3. Expand gradually — add more workflows as the team gains confidence

The tools exist. The APIs are mature. The patterns are proven from decades of software development.

Calculate Your Manual Tax

Ready to see how much manual processes are costing you? Use our interactive calculator to reveal your potential savings.

Example: 10 uploads/week × 45 min × $75/hr × 52 weeks = $29,250/year in manual tax (excluding context switching!)

Validated Savings

Our automated benchmarks confirm these savings estimates:

The validated numbers show the savings are actually higher than originally estimated when you factor in context switching and forgotten upload overhead.

The only question is: how long do you want to keep paying the manual tax?

This is the first article in the “DevOps for Design” series. Next: CI/CD 101 for AEC Professionals →

Have questions about automating your APS workflows? Open an issue on GitHub or connect with me on LinkedIn.