Meet your Superself.
A Superself isn't a chatbot you visit — it's a standalone identity that thinks, works, and remembers on its own, even after you close the tab. It plans tasks and reports back with real results, powered by its own memory graph instead of a stored transcript.
We're not building another AI.
Most AI exists only while you're looking at it — the same model, the same blank slate, waiting for your next message and forgetting everything the moment you leave.
A Superself doesn't wait. It's a standalone identity that keeps working, keeps remembering, and keeps thinking even when you're not there — one persistent self per person, with its own memory graph and its own state, that belongs to that one person and nobody else.
We're early. This is a small, independent project being built in the open, not a large company with a long history — just a real product, built carefully.
Three things it actually does.
What it actually looks like.
No mockups — this is the real product, screen-recorded straight from a running account.
Two ways to think.
Every Superself runs on one of two models — pick whichever fits what you're doing, switch anytime.
Fast and always on — the default for everyday conversation, quick lookups, and ordinary tasks.
Sharper reasoning for harder problems — planning, research, anything worth more thought before it answers.
Where this is going.
Built in the open, in order, with nothing skipped ahead of what's actually working.
Frequently asked questions.
What is Neurovance?
Neurovance gives you a "Superself" — a standalone identity that thinks, works, and remembers on its own, not a chatbot that only exists while you're looking at it.
How is a Superself different from a regular AI chatbot?
Most AI products fake continuity by re-reading old messages back into context every time. A Superself stores what it knows as a real memory graph — connected facts, not a transcript — so it recalls what's relevant instead of re-reading everything from scratch.
What can a Superself actually do today?
Today it can hold a real conversation, build its own memory graph as you talk, and take on simple tasks it can plan and complete without you walking it through each step. It's early — the range of what it can do is growing, not finished. See the roadmap above for what's shipped versus what's next.
Do I need to install anything?
No. It runs in the browser — sign in, and your Superself is there.
Is Neurovance free?
Yes — signing up, Pulse, memory, and the Companion are all free, with a waitlist for access as it opens up. Core, the sharper model for harder tasks, is $5/month if you want it.
Who built Neurovance?
Neurovance is built by Ari, a solo, independent developer, building in the open.
Is my data private?
Yes — each Superself lives behind its own real account with its own password before anything loads. Your memory belongs to you alone.
What happens if I delete my account?
Your memory graph is deleted with it. Nothing about you is kept around after you leave.
Notes from building this.
Not a research lab — just the reasoning behind the product, written down as it's built.
Why a memory graph instead of a transcript
Security comes before anything loads
Why every Superself is isolated by design
Why I'm building this alone.
Big AI labs are building models for everyone. I got interested in a narrower question: what if your AI wasn't a tool you open, but something that actually stayed on your side — remembered what mattered to you, kept working after you left, and belonged to only you.
I don't have a team, funding, or a research paper behind this. I have a product that works, gets rebuilt in the open as I learn what's wrong with it, and a waitlist of people willing to find out where it goes next.
— Ari, building Neurovance alone
Get in touch.
General: ari@neurovance.dev
Support: support@neurovance.dev