Questions and evidence

How can you use AI to prepare interview questions that enable competencies to be assessed on the basis of facts?

1250 € per session, excl. VAT

In short

This half-day session of 3 hours enables you to prepare a recruitment interview guide with support from generative AI. You link questions to the competencies required for the role, formulate neutral follow-up questions and test a framework of expected evidence, without delegating the assessment of applications to the tool.

Why this training?

An appealing question can elicit a rehearsed account rather than evidence of competence. Starting from situations actually encountered in the role helps distinguish between what the person claims, what they have done and what needs further exploration. AI can suggest wording and reveal duplication, but it can also introduce assumptions or irrelevant criteria. The recruiter retains responsibility for selecting and validating the questions and interpreting the answers. Continuing case study: The fictional company « Tressane Logistique » is recruiting someone to coordinate dispatches. Two managers must conduct comparable thirty-minute interviews. Using a fictional job description, participants explore three competencies: prioritising urgent issues, handling an incident and passing on reliable information. Transfer to the workplace: Within the following two weeks, adapt the guide to another role and have it reviewed by a qualified HR professional before use. After an interview, identify the questions that elicited vague answers and revise their follow-up questions, without passing any application data to the AI.

Objectives of this training

  • Translate three competencies required for the role into observable behaviours
  • Draft six behavioural or situational questions with AI assistance
  • Associate each question with a neutral follow-up question and expected evidence
  • Identify leading, intrusive or unrelated-to-the-role wording
  • Test the guide and measure its quality using a checking framework

Who is it for and what are the prerequisites?

Roles and tasks concerned

  • Recruiters preparing their interview guides
  • Operational managers taking part in interviews for their team
  • HR professionals harmonising practices across recruiters

Prior skills

  • Have already formulated and adjusted a simple request to a generative AI tool.
  • Know the stages of a recruitment interview and the duties of the role being studied.

What you will work on

  1. From the role to observable facts

    Objective: Define what the interview should enable you to observe. Distinction between duty, competency and behaviour; selection of critical situations; limitations of general impressions.

  2. Questions, follow-up questions and evidence

    Objective: Obtain wording that elicits specific examples. Behavioural and situational questions; follow-up questions on context, action and result; AI instruction specifying the role, constraints and format.

  3. Neutrality and comparability

    Objective: Revise the questions and their assessment reference points. Common core across candidates, distinction between absence of evidence and insufficient evidence, intrusive wording; legal points to check with your legal department.

  4. Simulation and adjustments

    Objective: Check the guide's usability in an interview. Time management, factual note-taking and analysis of vague answers.

What you take away

  • Matrix of three competencies and observable behaviours
  • Reusable AI instruction for generating and revising questions
  • Thirty-minute guide containing six questions, follow-up questions and expected evidence
  • Completed checking framework and correction log

Method and assessment

  • Short inputs followed by immediate practice using the continuing case study.
  • Individual work followed by work in pairs, with peer review of the outputs.
  • Guided group discussion: what works, what needs correcting and what remains a human decision.
  • Preparation: Prepare a fictional or anonymised job description, without CVs or candidates' contact details. Check access to a generative AI tool authorised by the organisation; a fictional case file is provided for the exercises.

Observable criteria: Each of the three competencies is explored through two questions explicitly linked to the role. ; Each question includes a neutral follow-up question and at least two factual elements to look for. ; No retained question presupposes an answer or concerns a personal characteristic unrelated to the role. ; The test keeps to thirty minutes and distinguishes between facts gathered, interpretations and missing information. ; The number of compliant questions before and after revision is recorded across the six questions..

Frequently asked questions

Can we work on a real recruitment process?

Yes, using only anonymised or fictional material, in tools authorised by the organisation.

Does the AI assess candidates?

No. No ranking or sensitive decisions are automated; assessment remains human.

Can the guide be used immediately?

It is ready to use, but must be validated by a qualified HR professional before actual use.

Must exactly the same follow-up questions be asked?

The common core remains consistent; follow-up questions clarify missing facts without changing the competencies being assessed.

Last updated: