Skip to content
OpenTrain AIOpenTrain AIOpenTrain AIDocs

Ask OpenTrain

Answers from the documentation, with sources.

What would you like to do with OpenTrain?

AI answers can be mistaken. Check the linked sources. Don’t include private account information.

Open app

OPENTRAIN AI DOCUMENTATION

Application screening quiz

Ask every applicant to complete a short course and screening quiz with their proposal, then review, grade, and rank the results before you hire.

An application screening quiz asks every AI trainer who applies to your job to complete a short course and quiz as part of their application. You see each applicant’s quiz standing and score next to their proposal, and you can rank applicants by screening score before you decide who to hire.

Use a screening quiz when you need comparable, graded evidence from every applicant, such as a check on your labeling guidelines, a domain knowledge test, or a sample task with a known answer.

Application screening quizAI interviewProject To-do quiz
Who takes itEvery applicant, after they submit a proposalApplicants, before they submit a proposalAI trainers you have already hired
What it isA published course with one or more graded quizzesA conversation with the AI interviewer, guided by your interview requirementsA published form or quiz assigned as a Project To-do
How you set it upThrough your AI agent, the OpenTrain CLI, the MCP server, or the APIIn the job wizardFrom the job’s Project To-dos
Where results appearThe Screening quiz column and summary in proposal reviewThe interview score and summary in proposal reviewForm results and the Project To-dos review queue

A screening quiz does not replace the AI interview. If the job requires the AI interview, applicants take both. If the quiz should be the only screening step, have your agent turn off the AI interview requirement for the job; the developer guide shows how.

To check the AI trainers you have already hired, use Forms and quizzes as Project To-dos instead.

You set up the screening link through the OpenTrain CLI, the MCP server, or the API. The simplest path is to ask your AI agent to do it. Linking a course from a job’s Training tab adds it as post-hire training for people you hire, not as an application screening quiz.

Write the quiz

Create a graded quiz with an answer key and, if you want pass and fail results, a pass rule. You can build it in Forms (see Forms and quizzes) or have your agent create it.

Put the quiz in a course and publish it

Add the quiz to a course as an assessment lesson. You can add short lessons before the quiz that explain the project or your guidelines. Build the course under Training in the employer dashboard sidebar, or have your agent author it. Publish the course; a screening course needs a published version.

Link the course to the job as its screening course

Your agent links the course to the job with the screening stage and sets how many quiz attempts each applicant gets. The default is one attempt. Applicants get the lower of this ceiling and the quiz’s own attempt limit. A job can have one active screening course at a time.

Publish the job

From now on, every new proposal on the job is enrolled in the screening course automatically.

For example, you could ask your agent:

Before you publish job <job-id>, link my published course <course-id> as the
application screening course, with 2 quiz attempts per applicant.
Then confirm the link and publish the job.

The developer guide has the complete CLI and MCP steps your agent follows.

  1. The applicant completes the proposal flow, including the AI interview if the job requires it, and submits the proposal.
  2. OpenTrain then asks them to complete the screening course, with a Start course button.
  3. They work through the lessons and the quiz at their own pace. Their progress is saved on the proposal, so they can leave and come back from the job page or from the proposals list on their dashboard.
  4. The course shows how many attempts each quiz allows.
  5. When they submit, their answers go to you for review.

Applicants see their score after submitting only if the quiz releases results immediately. If the quiz holds results for manual release, applicants see their score only after you release it. See Complete a screening course for the AI trainer view.

When a job has an active screening quiz, the job’s Proposals tab adds a Screening quiz column. Each applicant shows one of these standings, with the score once grading is finished:

StandingMeaning
Not enrolledThe applicant was never asked to take the quiz, usually because they applied before the course was linked.
Not startedEnrolled, but has not started the course.
In progressStarted, but has not submitted.
Awaiting gradingSubmitted. Some answers need manual grading before a score is available.
Passed or FailedFully graded against the quiz’s pass rule.
GradedFully graded; the quiz has no pass rule.

To compare applicants:

  • Sort the list by Highest screening quiz score or Lowest screening quiz score.
  • Filter by Passed, Failed, Awaiting grading, or Not completed.
  • Open a proposal to see the Screening quiz summary with the score, points, and submission date. The summary also notes when a result has not been released to the applicant yet, or when the applicant took an older version of the quiz.

Manual grading, result release, and final pass or fail decisions for a screening quiz run through the CLI, the MCP server, or the API. Your agent can also rank every proposal on the job by screening score in one request.

A screening result never hires or declines anyone. You still make the hiring decision on the proposal, as described in Review proposals.

If proposals arrived before you linked the screening course, those applicants show Not enrolled in the Screening quiz column. Your agent can find every proposal without an enrollment and enroll each one individually. After an applicant is enrolled, the screening course appears on their proposal for them to complete.

If an enrollment did not complete when a proposal was submitted, the same step repairs it.

If you publish a new version of the screening course while the job is live, new applicants get the new version. Applicants who were already enrolled keep the version they started. Their summary notes that they took an older version, because scores from different versions may not compare directly.