Zway.ai
Book a demo

Careers

The human layer is the product.

Zway hires operators, not marketers. An operator runs a small number of client accounts, publishes daily from accounts the client owns, and takes public responsibility for the work. We hire for judgement under scrutiny rather than for output speed, and we train on platform norms before anyone touches a live account.

Last reviewed 27 August 2026

Open roles

Four operator roles.

These are the jobs that do the work clients pay for. They are described here in full detail deliberately, because the most doubted claim we make is that the human layer is real.

  • Growth operator

    Owns a small number of client accounts end to end.

    The day
    Work the overnight queue first: drafts waiting, threads worth answering, anything the monitor flagged. Ship the high-value work before the client's morning. Afternoons are planning, client questions, and the work that needs thought rather than speed.
    What we train
    Platform norms per channel, the client's five documents, escalation judgement, and how to write a correction that a community respects.
    Brands carried
    A small number of brands, deliberately. Standing does not split.

  • Community operator

    Holds real standing in the places that ignore new accounts.

    The day
    Participate before promoting. Most of the day is answering questions in subreddits, Discords, and forums with no product mention at all. When a client fit is genuine, recommend it and disclose the relationship.
    What we train
    Per-community rules and unwritten norms, disclosure practice, de-escalation, and recognising the threads that are traps.
    Brands carried
    Fewer brands than a growth operator, because the accounts are personal.

  • SEO operator

    Ships technical fixes and owns AI-search visibility.

    The day
    Publish briefs, open pull requests against client repositories for technical fixes, and track which prompts each client is cited for in AI assistants. Argue with engineers occasionally, politely.
    What we train
    Technical SEO, working in someone else's codebase safely, schema, and generative engine optimization.
    Brands carried
    More brands than a growth operator, since the work is less conversational.

  • Creative operator

    Runs creator relationships past the first message.

    The day
    Outreach, negotiation, brief writing, and chasing. The work everyone assumes is automated and nobody has automated, because it is a relationship and it behaves like one.
    What we train
    Negotiation, creator economics, brief writing, and knowing when a partnership is not worth it.
    Brands carried
    Scales with campaign count rather than brand count.

How we hire

Judgement over volume.

  • What we hire for

    Judgement under public scrutiny. We can teach a platform in a fortnight. We cannot teach someone to read a room, concede a fair point, or resist replying when replying is the wrong move.

  • What we do not hire for

    Output volume. An operator who ships more by thinking less is a liability in this model, because the thing they damage is the account, and the account is the asset.

  • How we work

    Remote, asynchronous where possible, with real handover between timezones. Operators are told explicitly that escalating something is never penalised and that missing a deadline is cheaper than a public mistake.

If you are a founder rather than a candidate and you came here to check whether the operators exist, that is exactly why this page is this detailed. The full operating model is on how it works.

Questions candidates ask

Is this a writing job?
Partly, but not mainly. The first draft usually comes from the AI layer. Your job is to decide whether it should exist, make it true and specific, and then own it in public. Most of the skill is judgement about what to publish and how to handle what comes back.
Do I need marketing experience?
No. Some of our strongest operators came from support, community management, and journalism. What matters is that you have been accountable in public before, and that you can be corrected without becoming defensive about it.
Will I be asked to run fake accounts?
No, and if you are ever asked to, that is a mistake you should escalate immediately. We post from accounts the client owns and authorises, or from your own established account with disclosure. Fake personas and astroturfing are prohibited, and that is not negotiable for commercial reasons as much as ethical ones.
How many brands would I carry?
A small number, and fewer than you would carry at an agency. The exact figure depends on the role and the channels in scope. If we ever raise it to the point where the work becomes generic, we have broken the thing that makes the product work.
How do I apply?
Email hello@zway.ai with the role in the subject line, a short note about a time you were wrong in public and how you handled it, and anything you have written or shipped that you are willing to stand behind. A CV is optional and we read it last.

Apply.

Email hello@zway.ai with the role in the subject line. Tell us about a time you were wrong in public and what you did about it.

Email usRead the ethics policy first