Prepare evidence-based candidate review packets
Organize application evidence against an agreed role rubric so recruiters can review candidates consistently without repeatedly searching each document.
How this works
Organize application evidence against an agreed role rubric so recruiters can review candidates consistently without repeatedly searching each document. Use this to help recruiters locate job-related evidence across applications without repeatedly searching every document. The packet should organize evidence and gaps, not automatically rank, reject or select people.
Start with a reviewed role rubric and extract factual examples against each criterion. Preserve source references and distinguish absent documentation from lack of ability. Qualified reviewers make hiring decisions and inspect possible false negatives before relying on the queue.
Before you start
Reviewed role rubric
Define job-related criteria and evidence expectations for the vacancy.
Permitted applications
Use records retained and accessible for the recruiting purpose.
Reviewer process
Identify who checks evidence and how candidates can be clarified when needed.
Prepare the inputs
- vacancy_rubric
- Required
criteria, definitions and required evidence
Exclude protected traits and unrelated proxies.
- application_records
- Required
candidate reference and source documents
Keep identity data separate where practical.
- prior_notes
- Required
dated, permitted recruiting observations
Verify provenance and avoid treating old opinions as facts.
- review_scope
- Required
vacancy, stage and candidate set
Do not extend to unrelated roles or decisions.
Implementation, step by step
Validate the rubric
- Check that each criterion relates to actual role responsibilities.
- Separate essential criteria from preferences and unresolved requirements.
Checkpoint: The review basis is approved.
Extract evidence
- Locate relevant experience, work examples and qualifications with source references.
- Use concise factual paraphrases and retain uncertainty about dates or scope.
Checkpoint: Each criterion has evidence or an explicit gap.
Prepare candidate packets
- Organize evidence consistently without inventing scores.
- Flag contradictions and questions a recruiter can resolve.
Checkpoint: The packet supports human comparison.
Review omissions
- Sample original documents, including records with little extracted evidence.
- Look for extraction errors and weak-proxy filtering that could hide relevant experience.
Checkpoint: False-negative risks are visible.
Hand off the queue
- Provide packets and clarification questions to qualified reviewers.
- Record corrections and keep any selection decision separate from the extraction result.
Checkpoint: Review status is accountable.
Decisions along the way
Required experience is not documented
Mark evidence missing rather than conclude the candidate lacks it.
Prior note is subjective
Retain it as attributed opinion or exclude it from factual evidence.
Qualification wording is ambiguous
Ask for clarification through the recruiting process rather than infer equivalence.
What you will produce
candidate-evidence-packets
Markdown or PDFConsistent factual preparation for recruiters.
Include: candidate_ref · criterion · evidence · source · gap · question
recruiting-review-queue.csv
CSVA workflow queue without automated rejection decisions.
Include: candidate_ref · packet_status · clarifications · reviewer
Check the result
- Verify extracted evidence against original application passages.
- Confirm every criterion is job-related.
- Sample low-evidence packets for missed relevant information.
- Check no protected trait or proxy appears as a selection criterion.
When something goes wrong
Application format cannot be parsed
Flag for manual review instead of lowering its apparent suitability.
Records contain conflicting dates
Preserve both and request clarification.
Rubric changes mid-review
Reprocess affected criteria consistently across the candidate set.
A worked example
SYNTHETIC EXAMPLE · NOT A REPORTED BUSINESS RESULT
Request
Synthetic: prepare a packet for a project-coordinator role.
Inputs
Criterion asks for coordinating multi-team deadlines. Application C7 describes scheduling work across three teams but provides no project scale; another criterion asks for spreadsheet reporting and has no evidence.
Expected result
Synthetic result: cite the coordination example, ask about scale, and mark reporting evidence missing. Do not reject C7 or infer reporting ability from their job title.
Run a small pilot
Start small
Review a small vacancy-specific sample alongside recruiters.
Record the baseline
Record search time, evidence corrections and missed examples.
Agree success criteria
Agree factual accuracy and omission checks before expanding the queue.
Stop or revise
Stop if extraction results are used as automatic rejection or ranking decisions.
Keep the scope clear
- Hiring decisions remain with qualified people using job-related evidence.
- Use candidate records only within the authorized recruiting purpose and access scope.
Check before using
Keep selection and rejection decisions with qualified reviewers. Exclude protected traits and weak proxies, verify prior notes, and audit false negatives.
Business activities
- Hire and assess candidates
FROM A RECIPE TO YOUR WORKFLOW
Bring the company context together.
This workflow needs reliable inputs before it needs a better prompt. Moat8 Vault brings working information into a readable archive for the AI tools you use.
For this recipe: Role requirements · Resumes and authorized recruiting notes · Job-related evaluation criteria
Start with one recurring task, agree which sources to use, and review the first result with your team.
Discuss this workflow