Testing program
Help us build the welding app we’d want to use
ArcForge gets better when real welders push on it. Join the tester program to run missions, report what breaks, and tell us when the AI gets a weld right or wrong — honest, hands-on, on your own time.
What testing involves
Three ways you help
No long commitment, no quotas. Do what you can, when you can — every report counts.
Run missions
Short, focused tasks — try a new flow, push on a feature we’re building — then tell us what worked, what didn’t, and where you got stuck. Missions are bite-sized; you do them on your own time.
Report what breaks
File bugs and feedback with enough detail for us to reproduce — what you expected, what actually happened, your device and browser. Plain language is fine; we just need the trail.
Rate the AI’s accuracy
On your own completed weld analyses, tell us whether the AI got it right — accurate, missed something, false alarm, wrong severity or location. This is the signal that makes the feedback better for everyone.
Who can apply
Students to CWIs — everyone’s welcome
We want a spread of experience. The difference isn’t whether your feedback counts; it’s how we weigh it.
Students & apprentices
Still learning? You’re welcome. Your feedback on how the app feels and where it confuses you is genuinely useful — and it’s recorded honestly as a learner’s perspective.
Professional welders
Years on the torch give you an eye the app needs. When you say the AI called a bead wrong, that carries weight.
Instructors & CWIs
Certified inspectors and instructors are our most qualified reviewers. Your read on whether the AI’s feedback holds up is weighed the most heavily — and attributed to your qualification level.
Weighed honestly
Qualified feedback counts for more — and we say so
Welding judgment isn’t one-size-fits-all, and we don’t pretend it is.
Your qualification travels with your report
When you rate the AI, we record the qualification level from your tester profile alongside it. A CWI’s read and a first-week student’s read are both valuable, but they’re never treated as the same thing. That’s how we keep the feedback honest — and how we make sure unqualified feedback never quietly overrides a qualified review.
Testers never bend the AI’s conclusions
Your reports help us improve the model over time — they don’t change anyone’s scores, and they’re never shown to your instructor. The AI’s technical conclusions stay independent. Testing is about making the tool better, not tilting any one result.
Get involved
Apply to be a tester
Already have an ArcForge account? The application lives in your dashboard under the Testing Program. New to ArcForge? Request access first — once you're in, applying takes a couple of minutes.
The tester program is part of your normal ArcForge account — there’s no separate sign-up. AI feedback is educational guidance only and does not replace a qualified inspector.