الوظائف/المنصات/Handshake AI

How Handshake AI screens applicants

The steps Handshake AI asks applicants to complete before paid work, from its own pages (checked 20 October 2026), and what its 245 open postings say about tests and interviews.

خطوات الفرز

  1. Application
  2. Project match and interest email
  3. Decision and selection
  4. Project onboarding
  5. Training
  6. Project assessment (a quiz on the training, a practice task or a skills test, depending on the project)

Source: ai-support.joinhandshake.com

You sign up once for the Handshake AI Fellowship (profile, resume, work-authorisation questions), then wait to be matched to projects.

Assessment content

Assessments are project-specific and can be a knowledge quiz on the onboarding training materials, a practice task or sample submission reviewed by the team, or a written or skills-based evaluation; format, length and pass mark vary by project.

Source: ai-support.joinhandshake.com

Preparing for the assessment

  • Complete all training modules and onboarding materials first, set aside uninterrupted time, and follow each project's instructions carefully. (ai-support.joinhandshake.com)
  • AI tools: Fellows are prohibited from using LLMs or other AI-assisted tools to generate work product unless Handshake authorises it in writing. (ai-support.joinhandshake.com)

النتائج وإعادة المحاولة

  • There may be a processing period before results, and you are notified once a decision is made. Failing means you did not meet the project's minimum score or quality threshold; your profile stays active for other projects. (ai-support.joinhandshake.com)
  • Retakes: Retakes are decided per project and not guaranteed; requests are generally considered only for a documented error or technical issue, not to improve a score. (ai-support.joinhandshake.com)

Screening in open postings

  • 97% of Handshake AI's open roles (237) mention an ID check
  • 2% of Handshake AI's open roles (4) mention a skills test

From the postings' own text; a posting that doesn't mention a step may still use it.

Checks before paid work

  • ID check: Government-issued ID (such as a driver's licence or passport) verified through Persona during account creation
  • Paid onboarding: Some projects pay for onboarding time and others do not, as set in each project's Project Terms
  • Confidentiality: The Contractor Agreement includes mutual confidentiality obligations

من يمكنه التقدّم

US persons with work authorisation Handshake can support and an SSN or ITIN; international participation is invite-only. F-1 STEM OPT, J-1, H-1B and TN holders are not supported.

Handshake AI engages people as: independent contractor, not an employee.

Apply

See Handshake AI's 245 open roles · Join Handshake AI (referral link; it never changes your pay. More)

Facts from Handshake AI's own pages, checked 20 October 2026. Platforms change their process; their pages are the final word. See also how to spot a fake offer.

المزيد عن Handshake AI

Handshake AI screening: common questions

What does Handshake AI's screening involve?

Handshake AI's own pages list these steps: Application; Project match and interest email; Decision and selection; Project onboarding; Training; Project assessment (a quiz on the training, a practice task or a skills test, depending on the project). Steps can differ by project.

Does Handshake AI check ID?

Government-issued ID (such as a driver's licence or passport) verified through Persona during account creation

What does Handshake AI's assessment test?

Assessments are project-specific and can be a knowledge quiz on the onboarding training materials, a practice task or sample submission reviewed by the team, or a written or skills-based evaluation; format, length and pass mark vary by project.

Can I use AI tools in Handshake AI's assessment?

Fellows are prohibited from using LLMs or other AI-assisted tools to generate work product unless Handshake authorises it in writing.

Can I retake Handshake AI's assessment?

Retakes are decided per project and not guaranteed; requests are generally considered only for a documented error or technical issue, not to improve a score.