Fixed-scope offers · 02/03
Pilot on your data
One process, one metric, your real data. We build the model or agent, test it against your history and run it in shadow mode next to your people. At the end you decide with measured results: go live, iterate or stop.
In short
From
€10,500net
- Duration
- 6 to 8 weeks
- Who leads it
- An AI engineer and a data engineer, with an architect overseeing production readiness from day one.
What you get
- A success metric and a baseline measured before the pilot starts.
- A data pipeline for the use case, reusable in production.
- A working model or agent, evaluated against historical data and expert decisions.
- + 2
Where this leads
Measured results from projects of this kind, at our clients.
Who it is for
01
Companies that know which process they want to improve and how they measure losses.
02
Teams whose previous pilot worked in a notebook but never reached production.
03
Plants and operators that cannot afford a model that is wrong on day one.
What you get
01
A success metric and a baseline measured before the pilot starts.
02
A data pipeline for the use case, reusable in production.
03
A working model or agent, evaluated against historical data and expert decisions.
04
Shadow mode results: how often the system agreed with your people, and where it did not.
05
A production plan with costs, or a documented reason to stop.
How it runs
Step 01
Metric and baseline
What we improve and how it is measured today: scrap, complaints, hours per quote, downtime.
Step 02
Data pipeline
We connect the data the use case needs and give every record its context.
Step 03
Model or agent and evaluation
Built and tested against your history and expert decisions, with the threshold agreed up front.
Step 04
Shadow mode
The system works next to people, its suggestions are compared with their decisions, nothing changes in the process yet.
Step 05
Decision
Go live, iterate or stop, based on the numbers from shadow mode.
What can go wrong and how we handle it
- The model does not reach the agreed threshold.
- You see it in the numbers before go-live, not after. You get the analysis of why and what it would take to fix it.
- People do not trust the suggestions.
- Shadow mode exists for this: your people see where the system agrees with them and where it does not, and every suggestion shows what it is based on.
- Data turns out incomplete.
- We check data in the first weeks and tell you early, before most of the budget is spent.
What happens next
If the pilot meets the threshold, the production rollout reuses its pipeline and model. If it does not, you keep the code, the data pipeline and the findings.
Related case studies
Weld quality AI: 86% fewer defective parts
Global automotive parts manufacturer · Automotive, metal components
GenAI quoting agent: from 4 to 6 hours to 15 minutes
Manufacturing company · Manufacturing
Kit inspection with computer vision: 63% fewer complaints
Global manufacturer of industrial equipment · Industrial equipment manufacturing
Questions buyers ask
What if the pilot fails?
Then you learn it in weeks and at a fixed price, with the reasons documented. A stopped pilot with clear reasons is a valid outcome.
Does the pilot need to run in production?
In shadow mode it runs on real, current data, but it does not change the process. Nothing depends on it until you decide.
Can we start a pilot without the assessment?
Yes, if you already know the process, the metric and where the data is.
Let's talk about your project
Plan a pilot
Pick one problem.
Show us the data you have.
Let's agree a measurable goal.
Name the process and the data you have. An engineer, not a call centre, gets back to you.
We reply within one business day.