Candidate evaluation

Test how they actually perform.

A CV tells you what someone claims. A LivQuiz assessment tells you what they can do, under a clock, with the answers randomized so the result means something.

  • Real code questions in the language of the role
  • Randomized questions and options per candidate
  • Single use access links that expire on submit
Candidate assessment · Python
Question 4 of 1208:42 leftSingle use access
What does this function return when called with [3, 1, 2]?
def process(items):
    seen = set()
    return [x for x in sorted(items)
            if not (x in seen or seen.add(x))]
[1, 2, 3]
[3, 1, 2]
[]
Questions and options randomized per candidate
What changes

Interviews start from evidence.

20 min
To a ranked shortlist

Every applicant sits the same assessment, scored the same way, before anyone books a call.

Per question
Not one total

Score, time and topic for every question, so you know what to probe rather than starting cold.

Randomised
Per candidate

Question and option order differ per person, so a leaked screenshot is worth nothing.

Example scenario

Screening for a backend role.

A software team hiring one backend engineer
Illustrative example built from a typical hiring round
Example scenario
The problem

Twenty four applicants for one role. Screening calls were consuming a day a week and were not distinguishing between candidates who could code and candidates who could talk about coding.

What they did

Built a twelve question assessment mixing code reading with an explain aloud question, sent single use links to all applicants, and reviewed results before booking anything.

What changed

Nineteen completed. The median was 74%, and time per question showed which candidates had reasoned rather than guessed. Interviews started from the questions each person got wrong.

24 to 5
Applicants to interviews
19
Completed without an account
1 day
Screening time recovered each week
Results you can defend

A score you can put in front of a hiring panel.

Every attempt is timed by the server, not the browser, and every candidate sees a different order. When two people score differently, it's because they performed differently.

  • Server side timer that survives a reload or a dropped connection
  • A breakdown question by question, not just a total
  • Time per question, so you can see where they stalled
  • CSV export for your ATS or hiring record
Assessment results
Invited
24
Completed
19
Median
74%
Candidate A
18:04
92%Strong
Candidate B
23:51
68%Developing
Candidate C
25:00
41%Below bar
Breakdown by question, exportable to CSV
How it runs

From role brief to ranked shortlist.

Setting up an assessment takes minutes, and the same one can run for a single candidate or a whole intake.

01

Build

Write the questions yourself, or generate a draft from a role brief or job description and edit it.

02

Invite

Send single use access codes. No account needed on the candidate's side.

03

Assess

Randomized order, a server side clock, and a session that resumes exactly where it stopped.

04

Compare

Score, time, and question level detail for every candidate, side by side.

What you can ask

Questions that match the job.

Technical roles need more than four options and a guess. These formats are all available in the same assessment.

Code reading

Show a real function and ask what it returns, what breaks, or what the complexity is. Any language.

Randomization

Question order and option order shuffle per candidate, so a leaked screenshot is worth nothing.

Timed sections

Set a clock per question or per assessment. The timer is authoritative on the server.

Explain aloud

Ask the candidate to talk through their reasoning. The answer is transcribed and graded against a rubric.

Topic mapping

Every question maps to a competency, so a low score tells you which area to probe in the interview.

Export

Push results into your ATS or HRIS, or take the CSV and keep your own record.

Questions

Before you run one.

Do candidates need an account?

No. They open a single use link, enter their name, and start. Nothing to install and no signup wall between them and the assessment.

What stops someone retaking it until they pass?

Access codes are single use and expire on submit. Combined with randomized question and option order, a second attempt on a shared link is not possible.

What happens if their connection drops partway through?

The session is held on the server. They reopen the link and resume at the same question, with the clock reflecting the real elapsed time rather than resetting.

Which programming languages are supported?

Any language you can write a question in. Code is presented as authored, so Python, JavaScript, Java, Go, SQL, and anything else are all fine.

Hire on evidence

Stop interviewing people you could have screened in twenty minutes.

Build one assessment, send it to every applicant, and start the interview already knowing who can do the work.