We haven’t launched yet. We’re deep in the messy, unglamorous phase of testing everything like it’s the live system — breaking things, fixing things, making sure the beta is solid when it goes out the door.

And in the middle of all that, we rebuilt our infrastructure.

The question that started it

While talking with friends and people close to the project, the same question kept coming up: “You keep saying you care about privacy — but your website runs on Cloudflare, your database is on Cloudflare, your analytics are Cloudflare. Isn’t that all US infrastructure?”

They were right. And honestly, it was a fair challenge.

When we started building Voidcom, Cloudflare was the obvious choice. Cloudflare Pages is fast to set up, the CDN is excellent, and services like D1 and Web Analytics made it easy to ship quickly without managing servers. For a small team focused on building the actual product, it made sense.

But “it was convenient” isn’t a great answer when you’re building a communication platform that promises to respect its users’ privacy.

The gap between words and infrastructure

The EU-US Data Privacy Framework exists, and Cloudflare participates in it. Legally, data transfers were covered. But legal frameworks can change — and they have before. The US CLOUD Act gives authorities the ability to compel US companies to hand over data stored anywhere in the world. FISA Section 702 enables surveillance of non-US persons’ communications.

None of this means Cloudflare was doing anything wrong. But it means that relying on US-based infrastructure for a product that’s fundamentally about private communication creates a gap — between what we say and where the data actually lives.

We didn’t want that gap to exist when we asked people to trust us.

What we changed

We decided not to wait until after launch. If we’re going to ask beta users to trust Voidcom with their communication, the infrastructure should match the promise from day one.

Here’s what we moved:

Website hosting — from Cloudflare Pages to a self-hosted instance on Hetzner in Germany. We run our own deployment pipeline now. More work to maintain, but full control over where the data lives.

Content delivery — from Cloudflare’s CDN to Bunny CDN, an EU-based company headquartered in Slovenia. Fast edge delivery, DDoS protection, and no data leaving the European Union.

Website database — from Cloudflare D1 (US) to our own database on Hetzner in Germany. Newsletter subscribers, beta applications — all of it stays on our servers now.

Analytics — from Cloudflare Web Analytics to Plausible Analytics, self-hosted on our own infrastructure. No cookies, no personal data, no third-party involvement at all. The analytics data never leaves our server.

Code and CI/CD — we host our repositories on Codeberg, a non-profit, Germany-based Git hosting platform, and run our CI/CD pipelines through Woodpecker CI on our own self-hosted runners running on our own server hardware.

Installer downloads — served from Scaleway Object Storage in France. When you download Voidcom, the binary comes from EU infrastructure.

Newsletter emails — already handled by Proton Mail in Switzerland, covered by EU adequacy decisions. This one didn’t need to change.

Application servers — our backend is built to run on Hetzner in Germany. User accounts, messages, voice — all of it will be EU-hosted from day one.

This is just the beginning

We’re not claiming we’ve eliminated every non-EU dependency. Infrastructure is complex, and there are always more layers to examine — domain registrars, DNS, upstream network providers. We’ll keep evaluating and improving.

But moving the services that directly handle user data and website visitors to EU-based or self-hosted infrastructure is a meaningful step. It’s not a marketing checkbox — it’s a decision that costs us more time, more effort, and more operational complexity. We’re doing it because it’s consistent with what we believe.

Why now, not later

We could have launched the beta on Cloudflare and migrated later. It would have been easier. Nobody would have noticed.

But that’s exactly the kind of thinking we’re trying to avoid. “We’ll fix it later” is how products end up with privacy policies that don’t match reality. We’d rather do the hard thing now — in the middle of testing, under time pressure — and launch with infrastructure we’re proud of.

If we’re going to ask you to trust us with your conversations, your voice calls, your communities — we want to earn that trust before you sign up, not after.

Where things stand

We’ve updated our privacy policy to reflect all of these changes. It’s written in plain language, and it tells you exactly where your data goes, who processes it, and why.

We’re still in the final stretch before beta. Still testing. Still fixing things. But now when someone asks “where does my data actually live?” — we have an answer we’re comfortable with.

Apply for the beta and see for yourself.