Vauchi

Leave any platform without losing your people

July 10, 2026 · 7 min read

You don’t stay on a social network because its feed is good. You stay because the people are there — and leaving means losing them. In April 2026, the European Commission decided to keep it that way: it declined to extend the Digital Markets Act’s interoperability mandate — the rule that forces platforms to connect with their competitors — to social networking. Messaging apps like WhatsApp still have to open up; Facebook, Instagram, and TikTok do not, and there is no timeline for revisiting the question. The Commission’s reasoning, set out in its first DMA review report, was that there is little clear demand from users and businesses, and that the technical complexity is too high. The EFF called it what it is: a decision to keep EU users locked behind Big Tech’s gates.

We build a contact app, not a social network, so you might expect this decision not to concern us. It concerns us directly — because the thing the gates actually lock in is not content. It’s your relationships.

What the gates hold

Platform lock-in decomposes into two assets:

  1. The social graph — who you are connected to.
  2. The content layer — what flows over those connections, and who decides what you see.

Nearly the entire European regulatory debate is about the second layer. The DSA can audit recommender systems (the algorithms that decide what you see), ban manipulative design, and demand transparency, but researchers studying it conclude that the structural problem — feeds tuned to keep you scrolling, under the private control of a handful of companies — survives all of it. The most credible structural remedy on the table, letting users swap in independent recommender systems, explicitly depends on access to the first layer: the social graph. No graph access, no meaningful alternative.

And that is exactly the access the Commission just declined to mandate.

The graph is the moat — the reason leaving costs you people. Every “European alternative” and every federated network — the ones run by many independent operators, like Mastodon — runs into the same wall: your network doesn’t move with you.

Interoperability without permission

Here is the thing about your address book: it predates every platform, and it will outlive every platform. It is the one piece of social infrastructure that has never belonged to a gatekeeper — paper address books and phone directories, then SIM cards and phone contacts, then cloud accounts. The platforms’ great trick of the last twenty years was convincing everyone to rebuild it inside their walls, where it could be held hostage.

Vauchi is built on a simple inversion: the social graph belongs at the contact layer, held by the people in it, not by any platform — including us. The mechanics are deliberately simple: two people meet, scan each other’s QR codes, and each keeps a copy of the other’s contact card. That in-person exchange is the security anchor — you verified the person by recognizing them, not by trusting a platform. When meeting isn’t possible, two people can exchange through a shared link instead — and the card’s trust level, derived purely from how the exchange happened and editable by no one, honestly records the weaker verification. That way you always know which contacts you verified face to face and which you added remotely. From then on, the card is a direct, end-to-end encrypted channel between the two of you — meaning only you two can read it. Not the relay that carries it, and not us: we hold no key that opens it. It carries whatever its owner chooses to share: phone numbers, email, and — crucially — where to find them on any platform.

That last part is where the interoperability kicks in:

  • When someone moves — new messenger, new social network, new email, new country — they update their card once. Say Ana leaves WhatsApp for Signal: she edits one field, and everyone holding her card sees the new handle the next time their app syncs. The update travels end-to-end encrypted through a relay. The relay can’t read what changed, and because it routes by anonymous, daily-changing tokens, it can’t see who knows whom either. What it does still see: that an encrypted update passed through, and when (the threat model — exactly what we defend against, and what we don’t — is public for anyone who wants to verify). WhatsApp is never involved, or even aware.

  • Migration stops being a network event. Leaving a platform no longer means losing people; it means editing one field. The switching cost the Commission decided not to dismantle by mandate quietly deflates when the graph lives underneath the platforms instead of inside them.

  • No gatekeeper API is required. This kind of interoperability cannot be lobbied away, delayed for “technical complexity,” or declined for lack of “clear demand” — because it never needed permission in the first place.

There is a happier corollary. The same cards that make leaving cheap make arriving less lonely: as friends add their Signal or Mastodon handles, you can see — on your own device, visible to nobody else — which of your people are already there. The hardest part of joining a better platform was never installing it; it is the fear that nobody you know is on the other side. A contact layer answers that before you jump, one friend at a time — no campaign, no coordinated exodus, just individual choices becoming quietly visible to the people who already know each other. Lock-in works by hiding your options. A graph you hold yourself shows them — which may do more for the Signals and Mastodons of the world than any mandate would have.

What this doesn’t fix

We are not going to oversell this.

A contact layer only connects people who both use it. A regulation binds gatekeepers whether or not anyone switched yet; our approach helps you exactly as far as it spreads, one deliberate exchange at a time. That is slower. It is also the only route left standing. And there is intentionally no discovery layer: Vauchi connects people who already know each other; it doesn’t find you an audience.

It does nothing about feeds. Vauchi has no timeline, no recommender, no engagement metrics — by design, it is a tool you open when you need it, not a destination. If you care about fixing algorithmic curation (you should — the evidence connecting engagement optimization to polarization keeps piling up), that fight is at the content layer, and it still needs regulators, researchers, and better platforms.

And knowing where someone went is not the same as wanting to follow them there. A portable graph lowers the cost of keeping people when platforms churn; it doesn’t decide where the conversation happens.

The era this is built for

The platform landscape is fragmenting: X exoduses, Bluesky, Mastodon, and now EU-endorsed European platforms. Some of these will thrive; most will not; all of them ask you to rebuild your network inside their walls, one more time.

Every new wave makes a platform-neutral contact layer more valuable, because it is the only layer where all of those worlds can meet. The Commission has decided the gates stay closed for now. Fine. The more durable answer was never to ask the gatekeepers to open them — it is to hold your relationships somewhere no gate can enclose: with the people themselves.

That’s what we’re building — and building is the operative word: Vauchi is in active development, not yet released. If you want it when it ships, join the waitlist.


Text licensed CC BY 4.0 — quote and share freely, with attribution. ← All posts