State of Ethyx

If you have been watching Ethyx from the sidelines, here is the short version: we are in a closed beta, v1 is a real product sitting on a single server, and the next chapter is less “new chat features” than becoming a company that can actually see this through.

I wanted to write that down in one place, without a roadmap slide deck.

Where we are

Ethyx is in invite-only closed beta. That is not a tease. It is how we are testing the platform while I add people gradually—cautiously enough that we do not put too much strain on a system that was never designed for a sudden crowd.

v1 is already a product you can use. Multi-model streaming chat, conversation history, projects, search, usage visibility, encryption at rest, and the honesty work about what we protect and what we cannot. I am proud of that. Closed beta is not a holding pattern because v1 is unfinished. It is a holding pattern because the next problems are different.

I will keep adding people as we go. If you want in, join the waitlist—the form is at the bottom of this page—or contact us. I read those.

What v1 actually is

Under the hood, v1 is a single monolith. One app, one server, limited scalability. That is not a secret we are saving for a post-launch blog. That is the architecture we are releasing.

It is enough to run a careful beta. It is enough to find out whether people actually like this and want it. And if they do not—if there is no demand—then the scalability problem is no longer a problem. I would rather learn that now, with a box I can keep an eye on, than invent a distributed system for an empty room.

So yes: v1 is also proving the concept. Not as a prototype. As a working product with real users, asked a simple question: do you want this?

The turn we did not plan for

Ethyx started closer to a client you brought your own keys to. You talked to the models; we hid the SDK mess. That is a much simpler business than what we are now.

The move to a managed AI subscription—Ethyx runs the providers, you just use the app—needs plumbing we never designed for. Billing. Capacity. Operator keys. Abuse limits. Company-scale relationships with the same providers we used to just call on your behalf.

v1 still does the job it was built to do. That extra plumbing is v2.

Why the company step matters

The long-term privacy vision is to have Zero Data Retention agreements with providers—the kind of contract that is actually about what they keep, not a flag we send on a request and hope they honor.

We do not have those agreements today. I will not pretend otherwise. A no-retention flag is a request, not a guarantee. We have said that plainly in How honest is your AI chat app about your data? and Why Ethyx isn't end-to-end encrypted, and it is still true.

Those agreements are not something a passion project can sign. Providers negotiate them with operating companies: legal entity, bank account, live billing, someone on the other end of a contract. That is a large part of why v2 exists. Becoming a real company is not a branding exercise. It is how we get to sit at that table—and how we keep building this past the point where a nights-and-weekends project runs out of road.

What v2 is

v2 is the phase that turns Ethyx from a passion project into an operating company, while making the product fast and sturdy enough to open the gates later.

That means:

• The company work — the unglamorous stuff that makes contracts, public billing, and durable provider relationships possible
• Performance and scalability — so closed beta can become a wider opening without melting the box we have today
• Some new features — I am not going to list a wishlist. The product will keep moving. The point of v2 is the transformation underneath it

Closed beta stays in the foreground. I will keep testing the product, keep adding people when the system can take them, and keep working on the larger change in the background. The goal is still to open this up. We are just not going to pretend we can do that on the architecture we have, or without the company behind it.

Come try it

If you have been waiting for a polished public launch, that is still ahead. If you want to help decide whether this is worth scaling, the closed beta is the useful part.

Sign up. I will keep letting people in, a few at a time, while we figure out whether the demand is real—and, if it is, how to let everyone else through without breaking the thing they came for.