Signature offer · 8 weeks
AI proof of concept
A model trained on your data, measured on a test set, delivered in your dashboard — and an honest costing of industrialisation.
A measured model, delivery in your tools and a clear decision: industrialise or not, and at what cost.
Indicative duration 8 weeks, adjusted to the actual scope
An AI POC is useless if it stays a notebook. Ours delivers a measured model, delivery in your tools and a clear decision: industrialise or not, and at what cost.
Who it is for
- Business leaders or CIOs with a usable history: orders, invoices, tickets, measurements
- Organisations that want proof before investing
- Data teams that want a method transfer
How it unfolds
Weeks 1 and 2 · Scoping
Expected value, audit of the available data, target indicator, exclusion rules, success criterion.
Weeks 3 to 7 · Model and delivery
Data preparation, comparative training of several families of models, predictions on the target population, reporting dashboard.
Week 8 · Decision
Presentation of the measured results to the sponsor, costing of the industrialisation, decision.
What you get
- A scoping note
- Documented training and test sets
- The selected model and its limitations
- A prediction table integrated into your data model
- A dashboard comparing actual and predicted
- A costing and an industrialisation plan
Terms
- Fixed price, with the scope defined at the scoping stage
- Management in France, delivery at the Casablanca service centre
- Indicative duration, adjusted to the actual scope
- On quotation
Case studies
- Predicting on-time delivery, order by order — supply chain, international industrial equipment manufacturer
- Predicting the collection date of every customer invoice — finance, finance department of an industrial group
Have a project? Let's talk.
A first conversation to frame your needs, with no commitment, in Paris or Casablanca.