Become a design partner: early is a feature, not an apology
We're a young platform and we're not pretending otherwise. Here's exactly what design partners get, what we ask, and why early is the best seat.
This is the hundredth article in this series, and it's the one where we drop the analyst voice and talk plainly about where we are.
Rysh is early. Small team, young product, no wall of enterprise logos. If you've read this far into the series, you already know what we believe — that companies become AI-native by putting agents on real work, on one engine, where every action is observable, the risky ones are gated, and the logic and data are owned. And you know what we've shipped: the agentic terminal, autonomous agents from markdown files, humanoids on Slack, email, and website chat, browser automation, cloud pane sharing, loop budgets, approval gates, secret redaction — self-hostable, running on Claude with your own key.
What we don't have yet is you — the handful of companies whose real workflows will pressure-test that platform and shape what it becomes. That's what the design-partner program is. This article is the whole pitch, terms included, so you can decide in one read.
Why "early" is genuinely the best seat
The standard enterprise-software deal is: wait for maturity, buy the finished thing, adapt your workflow to it, get the same product as everyone else, and file feature requests into the void.
The design-partner deal inverts every clause. You arrive while the concrete is wet. The platform gets molded around your workflows — not retrofitted to them three years from now. The features you need jump the queue because you're one of a handful of voices, not ticket #4,182. And the expertise your team builds — how to run governed agents on real work — compounds from now, while your competitors are still writing AI-strategy memos.
There's a reason the best-fit platform teams at larger companies fight to be design partners for infrastructure they care about: influence over a platform's formative period is something you cannot buy later at any price.
Early has real costs too, and we'd rather name them than have you discover them: you will hit rough edges. Some features you want will be on the roadmap rather than in the build — for us, right now, that includes things like WhatsApp and phone channels for humanoids. What we owe you in exchange is honesty about what's live versus planned (a distinction this entire series has been careful about), fast fixes, and a direct line when something breaks. Early is a feature because both sides treat it like a working relationship, not a transaction.
What you get
Hands-on onboarding, on your actual workflow. We don't hand you docs and a login. We sit with your team, pick the first real workflow together, and get it running — agents defined, governance dialed in, budgets set. You leave the first sessions with something doing real work, not a sandbox.
A direct line to the founders. Not a support portal. The people who wrote the code answer your questions, watch how you use it, and fix what gets in your way. Feedback loops are measured in days.
Roadmap influence. Design partners are the primary input to what gets built next. If your workflow needs something, that need is argued about in our planning — by name.
Early-partner pricing. Locked-in preferential terms as the platform grows. We keep pricing honest and simple; joining early means the economics remember you were early.
Co-marketing, only if you want it. If things go well and you're willing, we'd love to tell the story together — case study, joint post, conference talk. Entirely optional. Quiet partners are equally welcome.
What we ask
This is a two-way commitment, so here's our side of the ledger — what makes someone a partner rather than a trial account:
Real usage. You put rysh on at least one genuine, recurring workflow — dev, ops, support, or automation — not a demo you poke at twice. The whole value of the program, for both of us, comes from real work hitting the platform.
Honest feedback, regularly. A short recurring check-in (we keep it tight) where you tell us what's working, what's broken, and what's missing. Blunt is better. You will not hurt our feelings; you will shape our roadmap.
Patience proportional to the stage. When you hit a rough edge, you tell us and give us a chance to fix it fast — rather than expecting the polish of a decade-old product from a platform you chose because it's early.
That's it. No long contract theater, no commitment to buy anything at the end.
Who this fits
You're a strong fit if you recognize yourself in one of these:
- An AI-forward engineering team (roughly 3–30 devs) already using Claude or similar tools, frustrated that your agents run outside your actual workflow — no shared visibility, no governance, no way to hand off context between people and agents.
- A company with an "adopt AI" mandate and no real plan — a founder, CTO, or head of ops who wants agents doing observable, governed work across the business instead of a scattering of disconnected AI subscriptions.
- A team that cares about ownership — you want the agent logic in your repo, your own API keys, and a self-hostable deployment inside your perimeter.
You're probably not a fit if you need a mature, hands-off product this quarter, or if your first requirement is a feature we've told you is roadmap. We'd rather say that here than waste your time on a call.
How it works
- Book a demo call. Thirty minutes. We show you the platform live — panes, agents, humanoids, sharing, governance — and hear what problem you'd point it at.
- Pick the first workflow together. One painful, recurring, real thing. We scope what "working" means for it.
- Onboard. We help you stand it up — on our cloud or your own infra — agents defined in your repo, approval and budget policy set the way your risk tolerance demands.
- Run it, and shape the platform. You use it on real work; we meet regularly; your friction becomes our backlog. Expand to the second workflow when the first has earned it.
The honest close
One hundred articles ago, this series opened with a claim: most companies are painting AI on, and the ones that matter next will be the ones that made agents part of how the company actually works — visibly, governedly, ownedly.
We built rysh because we believe that, and we're staking the company on it. What we need now is a small group of partners who want to prove it on real workflows — and who'd rather help shape the platform they'll depend on than rent whatever everyone else settles for.
We're early. That's exactly why the seat is valuable.
Become a design partner: rysh.ai/design-partner — or start from the beginning of the argument: What it actually means to be an AI-native company.