*This is a three-part piece. Part one is about where the idea came from. Part two is about the architecture I wanted to build. Part three is about the features that only became possible recently.
Part One: Twelve Years of Scars {#part-one}
There is a specific kind of frustration that only comes from building something really well inside someone else's constraints.
I spent twelve years — more or less from 35 to 45 — as the lead developer at e3 Systems Group in Mallorca, a specialist in connectivity services for superyachts. VSAT, Inmarsat FleetBroadband, cellular roaming, managed bandwidth. The kind of clients who expected perfection because they were paying tens of thousands of euros a month to stay online in the middle of the Atlantic.
Recurring billing for that world is not a simple thing.
You're dealing with monthly service fees, usage-based charges, data pooling across multiple vessels, currency conversions (yachts are registered in one country, operated in another, owned by a holding company in a third), and tax obligations that span multiple jurisdictions — all with clients whose accountants will scrutinise every line item. One wrong euro sign in a PDF and you get a three-hour phone call from a purchaser in Monaco. I cannot thank my colleagues (clients) Ana, Kasia, Lisa, ... enough for never exposing me to the client 'when it inevitably went wrong'.
I built CONNECT, e3's internal billing platform, from the ground up. Multi-currency. Multi-service. Multi-brand (e3 had several trading entities). Full audit trail. Integrated with Microsoft Dynamics Navision for accounting. Customer-facing iOS app at one point, so captains could see their invoices and pay online. API connections to VSAT billing systems that themselves had their own complexity.
By the time I was done, CONNECT had processed tens of millions of euros in recurring revenue. It worked. It really worked. The finance team could answer any question about any invoice instantly. The audit trail meant we could reconstruct exactly what happened to any billing record at any point in history.
And the whole time I was building it, there was a voice in the back of my head saying: this is brilliant, but it's built for e3. What if I could build the same thing — but properly multi-tenant, available to everyone, and without the constraints that came with working inside a single company?
What I Learned (the Expensive Way)
Twelve years of billing software teaches you things you can't learn from a textbook.
You learn that proration is a minefield. Changing a subscription mid-month sounds trivial until you're dealing with a customer who upgraded on the 17th, you're billing in arrears, there are usage charges that need to be apportioned, and the invoice needs to be legally correct under French VAT law. Every billing system I've ever seen has hacked its way around this. I wanted to get it right.
You learn that multi-currency is not just formatting. It's exchange rates at the time of transaction vs. time of payment, realised gains and losses, VAT bases that must be reported in the domestic currency regardless of what the client paid in, and VIES validation for intra-EU supply.
You learn that audit trails are the most important feature that nobody asks for — until the audit. The ability to prove, definitively, that a particular invoice line was generated by a specific rule at a specific point in time, and that no human ever touched it after generation, is worth more than almost any other capability in a billing system.
You learn that recurring revenue businesses have a specific rhythm that general-purpose accounting software simply doesn't understand. The concept of a subscription, a billing cycle, a renewal date, a proration event — these are first-class concepts in a billing system. In a general-purpose accounting tool, they're awkward workarounds.
And you learn, above all, that existing software is never quite right. Not because it's bad, but because it's designed for the average case, and the average case is never your case.
When I finally has the time in 2025, I had a clear idea of what I wanted to build. For ME!
Written by Kenneth Himschoot and published on April 3, 2026.