Your AI Forgets You the Moment You Close the Tab. Ours Doesn't.

There''s a piece of advice going around right now that''s genuinely good, as far as it goes: if you want an AI assistant that doesn''t sound like a stranger every time you open a new chat, write it three files. One describing who you are — your role, how you communicate, the frameworks you think in. One defining its voice — tone, values, the words it should never use. One giving it a name and a purpose, so it stops acting like a tool and starts acting like a team member.

It works, and it''s not a bad way to spend an afternoon. But sit with the shape of that solution for a second: you are now the one responsible for maintaining your AI''s memory of you. Every time your role changes, your working style shifts, or you learn something new about how you build best, that''s a file you have to remember exists and go update. The AI didn''t get smarter about you. You got better at describing yourself to it, in a text file, by hand, forever.

We think that''s still the wrong shape of solution

Not because the three-file idea is wrong — identity, voice, and purpose are exactly the right things for an AI to know. But if getting an AI to actually know you requires you to write and maintain the documentation on yourself, the AI hasn''t really gotten more personal. It''s gotten better at reading a form you fill out.

So when we built Cameron — Waymaker''s AI Cofounder — we asked a different question: what would it take for an AI to actually know who you are, how you think, and how it should show up for you, without asking you to write any of it down first?

Six things Cameron already knows — none of them from a file you wrote

How you work. Cameron reads your Builder DNA archetype — Visionary, Shipper, Architect, Hustler, Polymath, Craftsman, Operator, or Maverick — and adjusts its own tone, pace, and level of detail to match how that archetype actually needs to be talked to. Not a preference you set once. Read live, every conversation.

What you told it. Cameron proposes facts worth remembering as they come up in conversation — goals, constraints, working patterns — and asks before it keeps anything. Nothing is silently assumed. You can review everything it's saved about you at any time.

Where you are. Cameron knows the day, your energy window, and what you asked it last, so you're never re-explaining context you already gave it an hour ago.

What you disclosed. If you've told Cameron about an accommodation or your current mode, it honors that without you repeating yourself every session.

Your own words. One free-text note — "here's how to work with me" — that you write once and edit anytime. Not a template. Not a form. Whatever you actually want it to know, in your own words.

Files you hand it. If you do have something written down already — old notes, a strategy doc, a PDF — upload it straight into your Brain and Cameron reads it in. The difference is it's optional, not a prerequisite.

Same goal, different amount of homework

We're not arguing with the instinct behind the three-file approach — persistent identity beats generic output, full stop. We just don't think you should be the one maintaining the documentation on yourself so your AI can do its job. That work should be the product's job, running quietly in the background every time you show up — not yours, done once and then slowly going stale in a folder you forget to update.

---

Curious what Cameron already knows how to do for your specific way of building? Take the free Builder DNA assessment and meet your Cofounder. Take the quiz →