03 Features
History and retention
Data is available in live memory the moment it lands; as it ages, scheduled processes move it into the deep archive. Nothing expires.
Illustrative figures.
How it works
The data stays as it arrived
Most systems keep fine-grained data for a few months and then keep aggregates only. It is a choice you pay for years later, when the comparison is needed and the fine data is gone.
It lands and it shows
The sample enters live memory and is queryable the instant it was produced, without waiting for an overnight consolidation.
It ages and it moves
Scheduled processes move data from live memory into the deep archive. It is a transfer, not a reduction: no resolution is lost.
It stays queryable
A question about yesterday and a question about three years ago are written the same way. Where the data physically sits is not the querier's problem.
It is not rewritten
A recorded sample does not change. Corrections are new versions with their own date, so a comparison made today is still repeatable tomorrow.
What comes out
What it guarantees you
Retention is not a feature anyone notices until it is needed. Then it is the only one that matters.
Retention is decided onceWhere it is used
In every industry, no exceptions
The protocol changes, the way of working does not. It is why the same measurement layer sits under plants that have nothing in common.
Nearby features
The step before and the step after
How much history
do you actually need?
Tell us what checks you will have to face and over what period: granularity and retention follow from that, not the other way round.