About us
Kubli takes its name from the double: someone who looks exactly like another person but is someone else. It is a precise description of what we build.
Why it exists
In 2024 the European regulator wrote down something the synthetic data industry preferred not to hear: two datasets that look alike prove nothing about anonymity. Similarity metrics, which most of the market leans on, are not evidence. The evidence is trying to break the result and measuring whether you can. Kubli exists to do that part, which almost nobody does.
The name
Synthetic data done well behaves like the real thing: same patterns, same correlations. But it corresponds to no real person. The name says it in one word.
The mark
The K repeats three times, offset, at rising opacity. None of the three is the original and none is the copy: it is the guarantee the product sells, drawn instead of described. One colour at different depths, because three different colours would say the modules are separate things.
What we do
Companies have data they cannot use, locked up by privacy, regulation or the risk of recognising someone. Kubli makes it usable and attaches the measurement of the remaining risk, along with what that measurement does not cover. Kubli does not certify compliance: it provides the technical proof that whoever answers for it builds their decision on.
The four modules cover the four points where sensitive data blocks work: archives, flows, automated decisions, infrastructure. The principle is the same across all four: make usable what stays locked, without losing the ability to prove it.
How we work
We say what we don't do too
No claim is made without a number or a verifiable fact behind it. That goes for what the modules can do and for what they do not cover yet.
No certifications, real measurements
Kubli issues no compliance stamps. It delivers the measurement made and the document to back it up in front of whoever asks.
Your data stays where it is, when it matters
On Team and Enterprise the agent can run inside your network, and then the rows never leave it. What leaves is the count and the agent's status, not the data.
What we are not
Kubli holds no third-party certifications, issues no compliance stamps, and has no qualified electronic signature: what we call a signature is an integrity hash of the report. The thresholds we use are prudent conventions with EDPB rationale, not numbers set by law, and the report says so. Treated data does not automatically become anonymous under the GDPR: what we hand over is a defensible position, not a guarantee.
Two routes
If you have a problem to solve, start with a demo on your own case. If you are assessing us, as a fund or a procurement team, start with the confidential documentation.
If you have an archive sitting idle and want to know whether Kubli unlocks it, write to us: we reply within one working day.
Thirty minutes. You tell us which data is blocked and what you would do with it, and we show you the result on an example close to yours. If the case is not a fit we say so on the call.