Invisible

Who can join

• Worldwide remote is possible, but projects are matched to your profile, language, and location, so you are only shown and invited to projects that fit. Specific per-country availability shifts as projects open and close.
• Strong written English. It is a stated expectation across the trainer roles and central to the work (reviewing and writing model outputs).
• You work as an independent contractor, not an employee of Meridial or Invisible Technologies; you accept or decline each project.
• Two tracks: generalist roles marketed as 'no experience required' (reviewing/error-checking AI outputs), and specialist roles that require real domain expertise (e.g. law, STEM, finance, software engineering, linguistics). Most projects are specialist.
• Reliability and availability during a project's scheduled hours; ability to follow structured processes. Prior operations/QA or data-workflow experience is preferred and this is not aimed at complete beginners.
• No joining, membership, or application fee is charged to experts (standard for worker-side expert networks). Note: Meridial's own FAQ does not explicitly state a fee policy, so treat this as unconfirmed by official documentation.

Step by step

1. 1

   Browse projects and join the Meridial network

   Go to meridial.ai (the expert-network arm of Invisible Technologies), browse open opportunities on the projects page, and join the network. This is the entry point, not the invisibletech.ai corporate careers page.

2. 2

   Complete a detailed expert profile

   Fill out your profile with your specific domain expertise (law, STEM, finance, coding, linguistics, AI safety, creative writing, etc.). Matching is expertise-driven, so a vague profile means fewer or no invitations. State your niche precisely.

3. 3

   Profile review and matching

   Meridial reviews your expertise against current project needs. If there is a match they reach out with project details and next steps; if not, your profile stays active for future opportunities. Joining does not guarantee an engagement.

4. 4

   Pass the project assessment / screening

   For most AI Data Trainer projects you complete a skills assessment (and sometimes an interview or evaluation) before you are cleared. Reviewers describe it as time-consuming (up to several hours) and focused on logical reasoning and following instructions exactly.

5. 5

   Get assigned to a project and start paid work

   Once you pass, you are onboarded to a specific project or team and begin doing the work as a contractor. You only ever see projects matched to your profile, and project invitations spell out scope and fees before you accept.

Tips to get accepted

• Apply through meridial.ai, not the invisibletech.ai corporate careers page. The Meridial expert network (projects + profile) is the actual door for AI training contract work; the main careers page lists full-time corporate/engineering jobs.
• Make your profile expertise specific and narrow. Matching is done against project needs, so 'finance' or 'competitive programming in Python' beats 'generalist'. A precise niche is what triggers an invitation.
• On the assessment, answer strictly from the material provided and never inject outside knowledge or assumptions. This literal, instruction-following discipline is the exact skill they are screening for, and adding 'what you know' is the most-cited way people fail.
• Set aside a few uninterrupted, well-rested hours for the assessment. It is long and detail-heavy rather than tricky, and most avoidable failures are careless errors from rushing.
• Treat it as supplemental, invitation-based income and keep your profile active while you wait. Work comes in project waves, so a quiet period is normal and not a rejection.