OperationsAI ChatPrompt EnhancementCreative Libraries

Recruiters, HR teams, people ops, and hiring managers

Recruiting & HR

Prepare a clear role description, structured interview guide and blank job-related evidence rubric for human reviewers. Use AI to improve the materials, not to rank applicants, infer personal traits or make hiring and employment decisions.

Practical guidance

Know what you are making

A reviewed hiring-preparation packet tied to the role's actual work: job description, consistent questions, evidence prompts and interviewer instructions. It contains no applicant scoring, automated selection or promise that the process satisfies every applicable employment requirement.

Before you start

The work and necessary capabilities

Provide approved responsibilities, essential tasks, level and the skills needed to perform them. Distinguish required qualifications from preferences. Avoid shorthand such as culture fit or digital native that leaves reviewers to infer unrelated personal characteristics.

Approved terms and process rules

Supply confirmed location, schedule, compensation language, application steps and the organization's review requirements. Leave missing terms as placeholders. Have a qualified HR or legal professional determine which requirements apply to the role and location before publication.

Synthetic examples, not personnel records

Use fictional scenarios and sanitized role information for these templates. Do not provide resumes, medical information, protected characteristics, compensation histories or employee case files. The workflow does not need individual applicants to prepare a consistent interview packet.

Practical guidance

From input to reviewed output

  1. Define job-related outcomes

    Describe what successful work in the role looks like using concrete tasks and responsibilities. Ask chat to identify vague requirements and questions for the hiring manager. Keep preferences visibly separate so they do not become automatic exclusion rules.

    Checkpoint: Every proposed competency should connect to an actual task rather than a personal stereotype or an unexplained credential.

  2. Draft the role description

    Translate the approved brief into clear responsibilities, required capabilities and application information. Preserve the employer's actual terms. Ask for clearer language, not invented benefits, exaggerated flexibility or assurances about policies that have not been confirmed.

    Checkpoint: A human owner must verify every employment term and any required notice before the post is published.

  3. Prepare a consistent interview guide

    Draft questions around the approved competencies, with interviewer notes explaining the relevant evidence to listen for. Use a blank rubric with behavior-based descriptions. Keep it a preparation aid; do not ask the model to fill it using real candidate information.

    Checkpoint: The rubric should help a trained person record evidence consistently, not infer personality, disability or demographic characteristics.

  4. Rehearse with fictional scenarios

    Have HR and hiring managers inspect the questions and discuss how a hypothetical work sample relates to the job. Check clarity, relevance and the organization's accommodation process. Revise ambiguous wording before interviewing anyone.

    Checkpoint: A rehearsal is not validation of a selection test or evidence that a process is free of discriminatory impact.

  5. Approve and use through human-led processes

    Get qualified review of the packet, then use the organization's approved applicant and interview systems. Human interviewers record evidence and make decisions through established safeguards. Maintain a version record so candidates are not evaluated against unnoticed changes in the draft.

    Checkpoint: Do not let the generated document become an automated ranking, screening or employment-decision tool.

Technical details

What the workflow does—and does not do

OPM interview guidance + workflow boundary

Structure the materials, not a candidate ranking

OPM describes structured interviews using predetermined questions and consistent evaluation standards for job-related competencies. Here that informs preparation of a blank guide for human interviewers. It does not authorize AI scoring of candidates, personality inference or automated hiring decisions.

EEOC technical assistance; not legal advice

Job relevance requires qualified review

EEOC guidance discusses discrimination risks in employment tests and selection procedures and the employer's responsibility for appropriate validation and use. A model-written rubric is not evidence that those requirements are satisfied. Have qualified HR or legal professionals review the intended process for the role and jurisdiction.

App integration and data boundary

No applicant-system or personnel-data connection

This guide does not establish ATS or HRIS access, resume screening, employee monitoring or candidate messaging. Keep these prompts limited to authorized role information and fictional examples. Do not infer protected characteristics, health conditions or personal suitability from names, appearance, voice or other proxies.

Editorial safety boundary

Separate preparation from consequential decisions

A human-reviewed template can improve consistency of documentation, but it does not remove the need for appropriate accommodations, reviewer training and oversight. Leave ratings and decisions out of the chat workflow. Record the packet's version and approval separately from any real candidate assessment in the organization's approved systems.

Technical details

Choose by task, not by a universal ranking

These are editorial starting points, not measured comparisons. Check the exact model guide and current app access before using a model; linked catalogs may also contain historical entries.

Clear role and responsibility language

Evaluate a chat model on a sanitized role brief and ask it to distinguish requirements from preferences. Review whether the result preserves real terms without adding unsupported promises.

Tradeoff: A fluent job post is not confirmation of lawful wording, valid selection criteria or the employer's actual policies.

Explore OpenAI models

Interview-question preparation

Draft consistent questions and blank evidence fields around approved job tasks. Ask for missing information and ambiguity, not predictions about which applicant will perform best.

Tradeoff: A generated rubric is not a validated assessment. Qualified people must review its relevance and use it within the organization's process.

Compare Anthropic models

A concise clarity review

Use another available chat model to identify jargon, inconsistent terms and questions that do not map to the role. Keep the same approved brief for comparison.

Tradeoff: Model agreement does not establish fairness or compliance, and another model does not justify sharing confidential applicant or employee data.

Explore Mistral AI models

Practical guidance

Create, refine and review

Replace every bracketed placeholder with authorized context. These are original templates, not generated-output demonstrations or generation-tested recipes. Copying does not submit anything or incur a generation charge.

Create a role-centered job-description draft

Where to use it: Paste into AI Chat with an approved role brief, without applicant data.

Draft a job description for [ROLE AND LEVEL] using [APPROVED TASKS, NECESSARY SKILLS AND TERMS]. Separate required qualifications from preferences. Preserve confirmed compensation, location and schedule wording; leave unknowns as placeholders. Identify vague or unrelated requirements and questions for HR review. Do not invent benefits, infer personal traits or claim legal compliance. This is preparation material, not an applicant-screening tool.

Check: Have the real hiring manager and qualified HR reviewer approve the role requirements and published terms.

Refine a structured interview packet

Where to use it: Paste into AI Chat with the approved role description and fictional task scenarios.

Refine [INTERVIEW DRAFT] for [APPROVED JOB-RELATED COMPETENCIES]. Create consistent questions, evidence prompts and a blank behavior-based rubric for human interviewers. Use only [FICTIONAL WORK SCENARIOS], not real candidate records. Explain how each question relates to an actual task. Flag ambiguous criteria and accommodation-process questions for HR. Do not score, rank, reject or select applicants, or infer sensitive personal characteristics.

Check: Confirm the packet remains blank and that qualified reviewers evaluate its job relevance before use.

Review the preparation materials

Where to use it: Paste into AI Chat with the draft packet and approved policy excerpts only.

Review [HIRING-PREPARATION PACKET] against [ROLE TASKS], [APPROVED TERMS] and [AUTHORIZED POLICY EXCERPTS]. Identify unsupported promises, inconsistent criteria, irrelevant questions and unclear instructions. Cite the exact passage and propose a narrower correction. List questions for a qualified HR or legal reviewer. Do not analyze real candidates, assess personality, automate employment decisions or declare this process legally compliant.

Check: Treat the output as a drafting review, not a substitute for professional assessment of employment obligations.

Illustrative example

What a useful handoff could contain

An editorial example of structure, not a generated result, customer case study or measured outcome.

Illustrative role definition

For a fictional support-operations coordinator, the packet describes maintaining request records and explaining handoffs. Required capabilities refer to those tasks. Compensation, schedule and policy terms remain owner-supplied fields rather than generated facts.

Blank interview and evidence guide

A fictional scenario asks how an incomplete internal request should be clarified. The blank rubric prompts the human interviewer to record the steps described and their relevance to the role. It contains no real applicant details, ranking or recommendation to hire.

Professional-review handoff

Track which role requirements and question sets HR reviewed, the approved version and unresolved process questions. Keep accommodation arrangements and real assessments in the organization's authorized process, outside the example document.

Technical details

Check the result before relying on it

Quality checklist

  • All responsibilities and criteria map to approved job tasks rather than stereotypes, proxies or vague personal fit.
  • Employment terms and process notices are checked by the responsible human owner and qualified reviewer.
  • The prompts and output contain no real applicant data, sensitive-trait inference, ranking or automated employment decision.
  • The packet remains a blank preparation aid, with consistent questions, version control and human-review requirements.

When the result misses the brief

The rubric asks for culture-fit impressions

Replace vague personality judgments with job-related tasks and observable evidence fields. Ask the qualified reviewer whether the criterion is appropriate instead of letting a model redefine it.

A polished description invents benefits

Supply an approved terms sheet and require unknowns to stay blank. Never infer compensation, flexibility or employment policies from another company's typical job posting.

Preparation drifts into applicant ranking

Stop using individual candidate records in this workflow. Return to role definitions, fictional examples and blank materials; consequential employment decisions belong in an appropriately governed human process.

Provider maximums, picker estimates and actual customer charges are different things. Read the pricing explanation.

Common questions

Before you start the workflow

Can I upload resumes and ask who to hire?

That is outside this guide. It is designed for role descriptions, interview preparation and blank documentation, not candidate ranking or hiring decisions. Do not provide applicant records to these templates.

Does a structured rubric guarantee fairness?

No. Structure can support a consistent process, but a generated rubric is not validated or guaranteed nondiscriminatory. Qualified professionals need to review the criteria, administration and applicable requirements.

Can this replace our HR or legal review?

No. AI drafts can help organize language and identify questions. Employment policies, assessment methods, accommodations and publication requirements need review by qualified people responsible for the real process.

Evidence and review

Sources & boundaries

The linked technical references were reviewed on . Documentation and app availability can change.

This is general preparation and documentation guidance, not legal advice, a validated hiring assessment or a compliance guarantee. Official sources support the attributed interview and selection-procedure principles. No applicant was assessed, no employment decision was made and no applicant or HR system was accessed.

Opening EZ Ai Assist does not prefill these templates or automatically run the workflow. Choose the relevant mode and review your input before submitting it.

Continue exploring

View all use cases