We registered the name before we wrote a line of code. That's backwards, and I'd do it again, because the name was the thesis: documents have moved in two great waves, and we wanted to build for the third.
Here's the argument.
The first wave was paper
For most of business history, a document was a physical object with authority. The letterhead at the top wasn't decoration — it was proof of origin, a wax seal that had learned typography. A company's stationery was designed once, printed in bulk, and guarded. If something arrived on it, it was official. The signature at the bottom turned a page into an instrument.
Paper had virtues we forget: one canonical copy, obvious provenance, possession as proof. It also, obviously, didn't scale. Retyping. Carbon paper. Couriers.
The second wave was files
Microsoft shipped Word in 1983 — as Multi-Tool Word, before the marketing department intervened. WordPerfect owned the late eighties, and the .doc conquered the nineties. In 1993 Adobe released PDF out of John Warnock's "Camelot" project, whose goal was a document that looked the same on any machine, anywhere. In 2006 Google bought Writely and documents moved into the browser.
The file wave made documents infinitely copyable and instantly sendable. Real progress; we're not romantic about paper.
But notice what got lost. The letterhead discipline — designed once, guarded, canonical — dissolved. Now the "letterhead" is a .docx template a designer made in 2021, and everyone who opens it can nudge the logo three pixels, "fix" a font, or leave last year's address in the footer. Files gave everyone a printing press and no press operator. You can see the result in any small company's outbox: proposal_v7_final_FINAL(2).docx, three letterhead variants in circulation, an invoice still carrying the old tax ID.
Files made documents cheap to produce and expensive to trust.
The false third wave
Around 2023, everyone's documents started life in a chat window. That felt like a revolution, and half of it was.
Drafting genuinely collapsed. What used to cost a freelancer an evening — a first-pass services agreement, a proposal skeleton — now costs a prompt. But watch what happens next, because it's the same everywhere: copy the wall of markdown out of the chat, paste it into Word or Docs, and spend forty minutes doing everything the AI never saw. Hunting for the letterhead file. Fixing what the paste did to the fonts. Replacing the company address the model invented. Checking it still ends on two pages. Then export, email, and hope somebody prints, signs, and scans.
AI made drafting free and left shipping exactly as slow as before. The gap between "AI text" and "a document my company can send" isn't a detail. It's most of the job.
We don't believe chat is the destination for documents. Chat is an interview format. The deliverable was never the conversation.
The actual third wave
The third wave, as we see it: the document drafts itself — into the real thing, not into a text box.
Concretely, that means the document knows things. It knows your company's facts — legal name, addresses, tax IDs, bank details — as data, never as text someone retypes. It wears your brand as a system: letterhead, footer, and typography composed once, inherited by every document. It interviews you for what's missing instead of hallucinating it. It lays itself out on true A4 or Letter pages, because page count should be a property of the document, not a print-preview surprise. And it carries its own workflow: review, approval, signature, audit trail.
In that world the agent does the drafting and the layout, and you do the two things software should never do for you: decide the terms, and take responsibility.
That's the wave we named the company after. Paper, files, and now this. Triplewave.
Why us
We're Droidsize, a small studio in India, and we send the same paperwork every small company sends: proposals to overseas clients, offer letters, NDAs, invoices. Our own documents were the embarrassment that started this. For a company that ships software, our outbox was held together by a Word template and vigilance.
So we made ourselves the first test. Before we let Triplewave near anyone else, we rebuilt Droidsize's real letterhead — and a client's — pixel-close, using only what's in the product. If a document tool can't produce its maker's own paperwork, it isn't done. That's still the acceptance bar for every design feature we ship.
Two beliefs sit behind the product, both arguable.
First: most business documents should not be designed by hand. Not "designed once and reused" — not designed at all, by most people, on most days. Design should be a system that documents inherit, with overrides for the exceptions. This sounds like taking creativity away. It's the opposite of what a growing company needs from its paperwork, which is for creativity to stop showing up in the margins of invoices.
Second: the word processor is dissolving. Not dying — separating into two tools. A canvas for the people who design documents, and an agent for the people who just need one. The forty-year compromise between those audiences — one toolbar, three hundred buttons — served neither.
Waves don't drain the ocean. Paper still exists; files will too, and Triplewave happily exports .docx for the counterparty who redlines in Word. But we think self-drafting documents on real, branded, signable pages are what most paperwork becomes.
That's the bet. The product is live at triplewave.co, free while in early access. This blog is where we'll write down what we learn along the way — about documents, the laws around them, and the waves we call wrong.