The browser first, everything else after.
Every product runs fully in the browser and as a PWA. Native apps only where a product genuinely needs them. Track and Chat on mobile, Meet for room systems.
Company
Why we're building this.
Europe has the people, the law and the data centres for good work software. We're building it: sixteen tools that belong together, with one contract, every provider named, and servers in Germany.
Anyone who buys compliance with bad software loses users to shadow IT.
Product principles
Five rules we hold ourselves to.
Zero need for explanation.
A guest joins a meeting without an account. A task takes three seconds to create. If a feature needs training, it's built wrong.
Familiar patterns, no retraining.
We don't invent new interaction patterns where people already have well-learned workflows. The difference comes from suite connections, compliance and German realities.
German realities built in.
State public holidays, the Working Hours Act, retention periods, co-determination, payroll exports. These aren't special requests, this is the market.
Every product viable on its own.
You have to be able to book just one product. The suite wins people over by pull, not by force.
Sequence
One hero per wave.
Sixteen products can't be built all at once, and don't need to be. Every wave has to be sellable on its own. This is a sequence for the products, not a cut in scope.
Wave 0
Foundation
Account, organisation, roles, directory, calendar, files, editor, notifications, admin, billing
Everything after this depends on it.
Wave 1
Meet, Chat, Notes, Recap
The meeting tool that writes the minutes.
Meet is the fastest entry product. Guests with no account spread it on their own. Recap is the reason to pay.
Wave 2
Flow, Track, Board
Plan, book, bill, and the time tracking obligation is taken care of.
Flow and Track share a data model and a buyer. Board extends the team offering with creation.
Wave 3
HR, Assist
The personnel file that knows who works when.
HR needs the directory, calendar and Track as data sources. Assist needs enough products to be useful.
Wave 4
Sheets, Enterprise expansion
The analyst that knows the whole company.
Needs the data from all previous waves; evidence and on-prem in parallel for regulated customers.
To be candid
What can go wrong.
A young company that hides its risks loses credibility the moment one of them happens. So here they are.
Breadth against depth
The biggest risk isn't competition, it's sixteen half-finished products. The remedy: stick strictly to the waves, finish building each product before the next one starts.
The brand name
Use of the EU emblem is regulated, and brands that suggest official EU affiliation are regularly challenged. The abstracted twelve-dot ring is deliberately not the emblem, but the risk still sits in the name. Under trademark review.
AI hosting
Recap, Assist and Sheets need language models. We don't have our own GPU capacity. All models come from Mistral AI in France, connected directly: Recap transcribes with Voxtral and summarises with Mistral. We don't use providers outside the EU for AI.
Migration
Customers will only come if they can bring their data with them. That's why the importers are a product, not an afterthought. They haven't been built yet; they're in preparation.
Frankfurt
Developed in
Hetzner
Servers in Germany
Open
Subprocessors with country, including the USA
