Open your laptop right now and count what's open. A dozen tabs. A couple of PDFs. A half-finished doc. A Slack thread you meant to come back to. Somewhere in there is the thing you were actually trying to figure out — and by tomorrow, most of it will be gone. Not deleted. Just scattered back into the noise, waiting to be re-found the hard way.
That's the problem we started with. Not "how do we add AI to a browser," but something more basic: valuable knowledge is fragmented across the tools we use every day — tabs, documents, emails, conversations, and our own memory — and nothing has been built to hold onto it. Every tool sees a slice of what you're working on. None of them see the whole thing. So the work of connecting it all falls on you, every single time you come back to a project.
We think that's a solvable problem, and we think it's a big one — big enough to be worth building real infrastructure for.
What Trace actually does
Trace continuously captures and organizes the knowledge from the pages, documents, and conversations you interact with, and turns it into a knowledge graph — a structured, connected map of what you've read and what it relates to, instead of a pile of open tabs and forgotten bookmarks.
That knowledge graph is yours. You can search it in plain language, ask an AI to reason over it with full context instead of a blank slate, and — when you choose to — share pieces of it with the people you work with. The same graph that helps you find that one article from three weeks ago is what lets an AI agent act on your behalf with the right context, instead of guessing.
Why we're starting with a browser
We're starting here because the browser is where most knowledge consumption already happens — more waking hours than any other piece of software. It's the one place positioned to see the whole picture instead of a fragment of it. So rather than build another app you have to remember to use, we built the capture layer into the tool you're already in.
But the browser is the starting point, not the destination. Over time, we want Trace to become the infrastructure that lets individuals and organizations build, maintain, and share their collective intelligence — the layer underneath, not another tab competing for attention.
On privacy, because it matters here more than most places
A tool that organizes everything you read is only worth trusting if you're in control of it. Trace is private by default: your data lives locally on your device, and nothing goes to an external server unless you explicitly choose a cloud AI provider. Sharing is something you opt into, not a default you have to opt out of. We think "remembers everything" is only a good thing if it's your memory, under your control — not something happening to you in the background.
Where we're headed
This is early. We're a browser today, and honest about the fact that the bigger infrastructure vision is still ahead of us, not behind us. But the thesis doesn't change: knowledge shouldn't live in a hundred disconnected places, and there should be one layer — private, yours, and legible to both you and the AI working alongside you — that holds it together.
If that's a problem you've felt too, download Trace and see what it's like to stop re-finding the things you already knew.