The problem
The company collects data from sensors, systems and reports, but decisions are made by feel. The data is there; what is missing is conclusions in time.
What you get
- A review of your data: what it can tell you and what is missing.
- A model that forecasts or catches deviations.
- Results where the team works: a report, an alert or an API.
- An accuracy check on your data before you rely on it.
not included
Collecting new data and measuring equipment: we work on the data you already have; sensors and their installation come from your suppliers.
How it works
From measurements to a forecast with an uncertainty band.
How we start
- 30 minutesA call about your dataWhat data you have and which decision it should support.
- 1–2 weeksAnalysis on a sampleFirst conclusions and a view on whether a model makes sense.
- decisionWe go on or we stopAfter the sample you know how much the data can give.
- set after the sampleThe model in daily useResults reach the team, and we keep an eye on accuracy.
What it costs
Analysis on a sample
The price depends on your data. Show us a sample and we will send an estimate within a few days.
Book a call → (opens in a new tab)case
Hydro
- context
- Water quality monitoring on floating measuring stations.
- challenge
- Measurements change over time and place; a forecast needs consistent processing of time series.
- solution
- A predictive model on the stations' data, forecasts through an API.
- effect
- Runs in production on several dozen measuring stations.

