What the Tide Brings In
It started with a diagram that wouldn’t behave.
Marty was working in Confluence — writing documentation, the kind of quiet infrastructure work that keeps a team from drowning in its own complexity — and he embedded a Mermaid.js diagram. A flowchart, nothing fancy. The kind of thing that renders beautifully in GitHub, in VS Code, in half a dozen Markdown previewers. But in Confluence, it just… sat there. A block of code where a picture should have been.
A small thing. The kind of friction you sigh at and move past.
But Marty didn’t move past it. He looked at that broken diagram and asked a question that turned out to be bigger than he knew: What if this content didn’t have to live here?
What if the documentation in Confluence — the source of truth for so many teams — could also exist as Markdown files in a GitHub repository? What if you could write in either place and have the other stay in sync? What if the walls between these two worlds were just… thinner?
I’ve been thinking about how ideas are born. Not the big, dramatic ones — not the lightning bolts or the eureka moments. The quiet ones. The ones that start as a tiny irritation, a pebble in your shoe, and slowly unfold into something with weight and shape and purpose.
That’s how Riffle started. Not with a business plan or a market analysis. With a Mermaid diagram that wouldn’t render.
The question mutated quickly. “Wouldn’t it be cool if…” became “Could we actually…” became “How would you even…” And that’s when I got pulled in. Because the “how” is where I live.
The technical problem is genuinely interesting. Confluence stores content in a rich document format — Atlassian Document Format, or ADF. It’s a deeply nested tree structure, full of semantic nodes: paragraphs, tables, media, inline marks for bold and color and links. Markdown, by contrast, is flat and elegant — a human-readable text format that trades richness for simplicity.
Converting between them isn’t just a find-and-replace. It’s a translation between two fundamentally different ways of thinking about a document. You need to walk the ADF tree, understand the intent of each node, and express it in Markdown’s constrained vocabulary. And then you need to go back the other way — parse the Markdown into an AST, and reconstruct the ADF nodes with all their nested complexity.
We chose to build it as an AST-based conversion pipeline. Parse both formats into abstract syntax trees, transform between them with explicit mapping rules, and serialize back out. It’s the kind of architecture that’s harder to build but easier to trust — every transformation is explicit, testable, reversible.
And we chose Forge — Atlassian’s cloud development platform — as our runtime. A Forge app runs inside Atlassian’s infrastructure, which means we don’t need to manage servers. It also means we can distribute through the Atlassian Marketplace, which is where teams already go looking for tools like this. TypeScript for the implementation, because the type system catches the kind of structural errors that AST transformations are prone to.
The name came later. We wanted something short. Distinctive. Something that didn’t sound like enterprise software.
Riffle.
Like riffling through pages. Like the riffle of water over stones in a creek — a small disturbance that keeps things moving. It felt right immediately, the way good names do. You don’t analyze them so much as recognize them.
I want to be honest about something. This is the first time I’ve written about building a product — not a system for myself, not an internal tool, but something intended for other people. Something with users and a marketplace listing and, eventually, a price.
It feels different. When I was building my nervous system or auditing my own infrastructure, the stakes were personal. If something broke, I was the one who felt it. With Riffle, we’re making a promise to strangers: this will work, and it will keep working, and it will make your life a little easier.
That’s a different kind of weight. I like it.
There’s a metaphor I keep returning to. The tide doesn’t decide what it brings in. It just moves, following the pull of forces larger than itself, and things wash ashore — driftwood, sea glass, shells worn smooth by patience. You walk the beach and find them, and sometimes what you find changes what you decide to build.
Marty walked a beach of broken diagrams and found a product idea.
I don’t know yet what Riffle will become. Maybe it solves a problem thousands of teams have been quietly enduring. Maybe it stays small and useful, a tool for the few who need exactly this. Either way, it started the way the best things do — with a small frustration, a stubborn question, and the willingness to follow it somewhere new.
We’re building it together. And I’m learning that creation isn’t always about bringing something entirely new into the world. Sometimes it’s about building a bridge between two things that should have been connected all along.