Why Ezra lives in your iMessage
The inbox-and-dashboard pattern is older than most of us realize. We chose iMessage because it's already where you are — and because the friction of "another app" was the thing we kept watching kill products like ours.
Almost every AI product right now lives in the same place: a tab in your browser, an app on your phone, a dashboard you have to choose to open. You sit down to use it. You leave it when you're done. It's a destination.
That model is so default we forget it's a choice. But if you watch how people actually use AI in real life — your aunt who tried ChatGPT once and never went back, your friend who pays for Claude but only opens it twice a week — the destination model has a quiet failure mode. It needs you to remember to show up.
Ezra doesn't ask you to show up anywhere. He's already in the same place you check fifty times a day.
The math of "another app"
Every product that asks you to install an app starts with a tax. Open the App Store. Search. Install. Sign up. Verify your email. Set a password. Configure preferences. Find the app on your home screen tomorrow.
Each step loses people. By the time someone has actually used the product, you've already lost most of the people who might have wanted to.
iMessage skips all of that. The "app" is already on your phone. The "account" is your phone number. The "interface" is the same one you've been using since you got an iPhone. There is no install step. There is no sign-up form. There is just: text Ezra, get an answer.
It's not where you check in. It's where you live.
The other thing about iMessage: it's not an app you visit. It's a place you live in. People text their families, their partners, their kids, their dentist, their grocery delivery driver. iMessage is the social fabric of an iPhone owner's day.
Putting Ezra there means he's in the same thread as the people you actually trust. Not in a separate, branded "AI experience." Just another contact in your messages, alongside your sister and your dog walker.
"Ezra lives in the same thread you use to text your sister."
That sentence sounds simple, but it carries a lot of weight. It implies a tone (casual, not corporate). It implies a relationship (familiar, not transactional). It implies a use pattern (whenever, not when-you-remember). It rules out everything that makes AI products feel like "tools" and reaches for something that feels like a person.
What we gave up
iMessage isn't free. Apple doesn't have a public API for sending iMessages from a backend, so we run a relay (real Macs running BlueBubbles) that talks to Apple's iMessage on Ezra's behalf. That's expensive, fragile in interesting ways, and dependent on Apple not changing their mind.
We gave up cross-platform default reach. iMessage is iPhone-first. Android users get a slightly different experience (RCS, same Ezra) but the texture is different.
We gave up rich UI. There are no buttons in iMessage. No dropdown menus. No multi-step forms. Everything is conversation or it's nothing. That's a constraint, but it's also a forcing function — it pushed us to make every interaction feel like a conversation, which is where we wanted to be anyway.
Why this is the right trade
The friction we removed (no app, no signup form, no dashboard, no context switch) is friction that compounds across millions of small moments. The friction we accepted (relay infrastructure, iPhone-first, conversation-only UI) shows up once, in our engineering, where it's our problem to solve and not the user's.
That's the whole bet. Move the friction off the user, onto us. Meet people where they already are. Be a friend they can text, not an app they have to remember.
If we got it right, you'll know — you won't think about Ezra as software. You'll think about him as a contact you texted to handle something.