Get started

Turn client requests into jobs with your AI assistant

When a client emails one or more job descriptions, paste the text into your AI assistant connected over MCP. It maps each description onto the fields of an Aptora job, validates them and creates one private draft per opening in your team. Nothing is published until you say so.

How it works

  1. Starting point

    A client sends two openings

    The email describes role, location, timeframe and requirements in prose, sometimes several openings in one message.

  2. Action

    Email into the assistant, with an instruction

    You paste the email into Claude, ChatGPT or Cursor and ask the assistant to create job drafts in Aptora. It reads the job format, validates each draft and creates it, one call per opening.

  3. Result

    Private drafts with detected skills

    Each job sits in your team as a draft: visible only to you, not on the job board and not in matching. Aptora detects the skills in the description; you mark them as required or optional.

  4. Next step

    Review and publish

    You go through the drafts under team jobs in the app, add missing skills and publish there or via publish_job. After that, you can search for talent for the job.

Example: two openings from one client email

Client email (excerpt)

Our platform team needs a Kubernetes engineer from 1 December: freelance, 6 months, 80% remote, based in Hamburg. Also permanent: a data engineer with Python and Airflow, starting in the first quarter.

Your instruction to the assistant

Create two job drafts in Aptora from this email; the company is Example Logistics Ltd. Validate the details first and don't publish anything yet.

Calls over MCP

  • get_job_schema
  • validate_job
  • create_job

Result in Aptora

  • Kubernetes engineer · Freelance · Hamburg, 80% remote · Draft
  • Data engineer · Permanent · Draft
  • Skills detected in both descriptions, ready for your team to review

Requirement: Any team workspace can create jobs. Automatic skill detection is included from Team Lite; without it, you assign skills yourself in the app before publishing. When you connect the assistant, you grant it the job permissions. Plans and pricing

One draft per opening

The assistant creates a separate job for every opening. Title, company and type of role, freelance or permanent, are required. Location, remote share, timeframe, day rate or annual salary and the full description are added when the email provides them. If something is missing, the field stays empty until you fill it in.

Validate without saving

validate_job checks a draft against the job format, names invalid fields and can preview the detected skills. Nothing is saved.

Skills before publishing

A job needs at least one skill before it can be published. Whether a skill is required or optional sets how much it counts in matching. Your classification, in the app or via classify_job_skill, stays in place even when detection runs again later.

Nothing goes out without you

create_job publishes nothing. A job only appears on the job board and in matching once you publish it: in the app or by explicitly asking your assistant to.

Frequently asked questions

Can I take over a job description without an AI assistant?

Yes. In the app's job editor you import a description as a PDF, screenshot, photo or text file. Aptora fills in the form; nothing is saved until you create the job. Importing is included from Team Lite.

What happens to several openings in one email?

Each opening becomes its own draft, one call per job. A daily limit per team prevents accidental bulk creation.

Which permissions does the assistant need?

The jobs permission to create and edit, plus the publish permission to publish. You choose both when the assistant signs in. You can revoke the connection in your settings.

Related

Get started

Decide per category. Necessary cookies cannot be switched off — without them sign-in and language do not work.

Keeps you signed in (Clerk), remembers your language and time zone so dates and texts render correctly, and records this cookie choice itself.

Stored for: session up to 12 months

Reach and usage measurement with Plausible Analytics, which we host ourselves on our own infrastructure. Cookie-free and with no cross-device recognition; no data is passed to third parties. It records page views, in the signed-in area without identifiers, and individual product events such as "signup completed" — no names, e-mail addresses, CV or job-posting content.

Stored for: nothing stored on your device

Measuring whether our ads on Google, Meta and LinkedIn work. If you arrive through an ad or a campaign link, a cookie keeps the campaign details from that link, the platform's click ID, the page you landed on and the domain of the referring website. If you register, we store this with your account and report to the platform you came from only the click ID, the event and its value — no e-mail address and no profile data. We load no scripts from Google, Meta or LinkedIn for this.

Stored for: cookie 90 days, in your account until it is deleted