A senior designer on retainer.
Faster, because AI drafts.

Enzo E, the studio's founder, is your designer and does final review on everything. AI accelerates the draft: faster turnaround, more on-system variations, early inconsistency checks. It never has the final say.

48 hours is the target for a standard requirement, from submission to a reviewed, on-system deliverable in code and Figma. Larger requests are scoped upfront.

Design-system maintenance and extension: components, tokens, patterns, and the documentation around them. Net-new product strategy or research is out of scope, and your designer will say so plainly if a request needs it.

Cancel anytime from the Stripe customer portal, no contract and no notice period. You keep everything already delivered. There is no pause option today, so a cancel and restart is the way to take a break.

Studio is $3,995 a month and Scale is $6,995 a month. Studio gives you one concurrent request slot, Scale gives you two. Both are month to month with no contract.

The number of concurrent request slots. Studio has one slot, Scale has two, so Scale keeps two requirements moving at the same time. Everything else, the designer, the review, the 48 hour target, is the same on both.

A slot is one requirement being worked at a time. When a deliverable is reviewed and handed back, the slot frees up and the next request from your queue moves into it. Studio has one slot, Scale has two.

No. Wraith is month to month, with no contract and no notice period. You cancel from the Stripe customer portal whenever you want.

Yes, from the Stripe customer portal, and it takes effect without a notice period. You keep everything already delivered. There is no pause option today, so a cancel and restart is how you take a break.

No, there is no pause option today. Cancel from the Stripe customer portal and restart when you are ready. Nothing already delivered is taken away in between.

You keep everything already delivered. The work lives in your own Figma file and your own code, so cancelling does not remove anything or lock you out of it.

You do. Deliverables land in your Figma file and your codebase, and you keep everything delivered, during the subscription and after it ends.

No. Your files, your requests and your design system are not used to train models, and nothing from your workspace is reused as context for another client's work. AI drafts against your system to do your work, and that is the only thing it is used for. Confidentiality is contractual, not just a policy: a mutual NDA can be signed before any work starts.

Yes, a mutual NDA at no extra cost, and it can be signed before any work starts. Confidentiality is the default either way: your requests, files and code are not shared outside the engagement, and nothing is published as case study or portfolio work without your written permission.

AI does the drafting, inside your design system, and a senior designer reviews and signs every deliverable. That split is what buys the speed: more on-system variations and early inconsistency checks, without AI having the final say.

The same named senior designer, on every deliverable. Nothing reaches you without that review. AI never signs off on its own output.

Unlimited revisions, for as long as the request holds its slot. If a deliverable is not right, you say what is off and it goes straight back into the same slot for another pass at no extra cost, rather than becoming a new request. The slot only frees up when you are happy, so the real limit is your own throughput, not a revision count.

Through the client portal. You write the requirement, it enters your queue, and it moves into a slot when one is free. Standard requirements target 48 hours from submission to a reviewed deliverable.

You set the order of your own queue. Whatever you put next is what moves into the slot when one frees up, so priorities are yours to change at any time.

Anyone you invite, at no extra cost. There is no seat limit: admins and members can submit and edit requests, viewers get read-only access, and the workspace owner is an admin by default. What is capped is concurrent slots, not headcount, so a bigger team simply shares the same queue.

It is async by default. Requests, questions, revisions and deliverables all live in the client portal, and you get a reply within one business day, so there are no standing meetings to attend. Your designer works US Eastern hours, and keeping the loop written rather than scheduled is part of what protects the 48 hour target.

You get portal access the day you subscribe and can submit your first request immediately. Before the 48 hour target applies, your designer needs access to your Figma file and your repo so deliverables land on-system rather than as flat mockups, which usually takes the first few days. After that, every standard requirement runs on the normal 48 hour target.

In the client portal, plus your own Figma file and your own code. Every deliverable arrives on-system in both, so there is no separate handoff step to translate a mockup into components.

Both. A standard deliverable is reviewed and on-system in code and in Figma, not a flat mockup you have to rebuild.

Yours. Components come back in the framework your codebase already uses, and tokens come back in the format your system already speaks, read from your real Figma variables rather than a parallel set invented on the side. React and TypeScript with CSS custom properties is the common case, but the deliverable conforms to your repo's conventions and your existing token names, not ours.

There is no separate handoff step. Work arrives as a pull request on your own repo alongside the updated Figma file, so your engineers review design work the same way they review any other change. Wraith's GitHub app can also run an automated design-token check on your pull requests, flagging hardcoded values that should reference a real token; it is informational and never blocks a merge.

WCAG 2.2 AA. Contrast, focus states, keyboard operability and sensible semantics are part of the review on every component, not a pass bolted on later. That is the standard the work is designed to rather than a certified audit, so if you need a formal audit or a VPAT, that is separate work with a specialist.

Wraith works on an existing design system: maintaining and extending its components, tokens, patterns, and documentation. If you do not have one yet, ask before subscribing, because building one from scratch is a different piece of work from what the subscription covers.

Net-new product strategy and research are out of scope. If a request needs either, your designer will say so plainly rather than quietly doing a thin version of it.

A subscription is month to month at $3,995 or $6,995, with no contract, no notice period and no hiring process. You get one senior designer's review on every deliverable, scoped to design-system work rather than the full remit of an in-house hire.

Scale's two concurrent slots are the ceiling, so if you consistently need more than two requirements moving at once, an in-house designer or a full agency is the honest answer, and your designer will say so rather than stretch the subscription past what it does well. Leaving costs you nothing: cancel from the Stripe customer portal and the work is already in your Figma file and your codebase, so there is nothing to migrate.

It gets scoped upfront instead of being promised in 48 hours. The 48 hour target covers a standard requirement; anything larger is agreed on before work starts, so you always know what is coming and when.