A working prototype on your real data, before you commit
The lowest-risk way to find out if AI can take over a process. We build a working version on your real cases, measure it against targets we agree up front, and give you a clear go or no-go.
LengthFixed price
Proof of concept Invoice matching
- MapProcess mapped, 180 real invoices collected, pass mark agreed: 95%
- BuildPrototype built and tested on the sample
- ProveYour team tries it. Results report and build plan
In short
What is an AI proof of concept?
An AI proof of concept is a small working version of an AI system, tested on a company's real data, built to show whether the idea works before the full project is paid for. Acropine starts with a discovery session to pick one process, then builds a fixed-price prototype and reports how accurate it was on your own data.
What’s included
- Process mapping
- One process mapped with the people who run it: steps, systems, volumes and exceptions.
- Agreed pass marks
- The accuracy, speed or cost the system must reach, agreed in writing before we build.
- A working prototype
- Built on your real past cases, including the messy ones, that your team can try.
- Results report
- How the prototype performed on your cases, where it failed, and why.
- Cost estimate
- What the full system would cost to build and to run each month.
- Build plan
- A fixed quote and timeline for production, if the numbers support it.
How it works
- 1
Map
Workshop with your team, collect sample cases, agree pass marks.
- 2
Build
Prototype built and tested against your cases as we go.
- 3
Prove
Your team tries it, and we share the results report and the build plan.
A good fit if
- You have one process in mind
- You want proof before a big budget
- You have past examples of the work to test on
- Someone on your team can review results with us
What you get
- A working prototype
- A results report on your data
- Running cost estimate
- A fixed quote for the full build
Common questions
What if the results aren’t good enough?
You still get the full results report. Often it shows a smaller part of the process that AI handles well. If it’s not worth building, we tell you plainly.
Is the price credited if we go ahead?
The proof of concept is quoted as a fixed price. Much of what we build carries straight into the production system, so little work is repeated.
What data do you need?
A sample of real past cases, such as tickets, calls, documents or enquiries, and read-only access to one system. We sign an NDA before anything is shared.