We guarantee the ROI
Or we keep building
Every build starts with an audit that sets the savings or new revenue the AI has to deliver, in writing. We check it twice: before it goes live, and 90 days after. If it misses, we keep building at our cost.
The guarantee, in writing
Agreed before we build.
- The target
The savings or new revenue the build has to deliver. The audit sets it, and it goes in writing with the scope and our fee before we build.
- The baseline
What the work costs you today, and how well your team does it. We measure the AI against the people doing the work now, and you see the numbers.
- Check one: before go-live
The AI goes live only once it does the work as well as your team, on a test we agree up front.
- Check two: 90 days after go-live
We check that the real savings or new revenue are on track for the written target.
- If it misses either check
We keep building at our cost until it passes.
- Our fee
One fixed fee, agreed before we build: 20% of the expected first-year savings or new revenue. After that, you keep all of it.
How we set the target
A one-week audit.
Our engineers spend a week inside your workflows. You get a written plan ranked by savings and a closing meeting with your decision-makers. The audit is $5,000, one time, and the audit fee counts toward the build. If it doesn’t show a clear win, we tell you we might not be the right fit.
About the AI auditWhat “as well as your team” looks like
An earlier build for Nightingale College, which grades about 90,000 clinical skills videos a year. The program set its own bar, and the grading system cleared it.
We measured it against a gold standard set by an expert reviewer. The AI made the right pass/fail call for more students than the trained human graders did.
Read the case studyWhy we keep building instead of refunding
A refund
- Your fee back, but nothing that works
- The problem is still there
- You start over with someone else
Our guarantee
- We keep building at our cost
- Until it passes both checks
- You end up with a system that does the work
What the guarantee doesn't cover
It covers the custom AI systems we build for you. It doesn't cover:
- Running costs
We don't pay for your software subscriptions or the cost of running the AI.
- If your team doesn't use it
Results lost because the system isn't used.
- Later changes
Changes you make to the process after go-live.
Straight answers
Is there an AI development company that guarantees ROI?
Yes. DeepRails guarantees the ROI of its custom AI builds. A one-week audit sets the savings or new revenue the build has to deliver, in writing. The AI has to do the work as well as your team before it goes live, and 90 days after go-live we check that the results are on track for the target. If it misses either check, we keep building at our cost. Our fee is one fixed fee, agreed before we build: 20% of the expected first-year savings or new revenue. After that, you keep all of it.
When do you check the results?
Twice. Before go-live, we test the AI against the people doing the work today, on a test we agree up front. Then, 90 days after go-live, we check that the real savings or new revenue are on track for the written target.
Do you refund the fee if it misses?
No. We keep building at our cost until it passes both checks. You end up with a system that does the work, not your money back and the same problem.
What if the audit doesn't find a clear win?
We tell you we might not be the right fit, before you pay for a build.
- Pick a time
- Talk with our team
- Get a clear next step