Playbooks·HR / People·Professional Services·HR03

Screening & Candidate Shortlisting

Extract evidence from candidate materials, generate summary cards, and produce a criteria-linked shortlist with risk flags.

GenAI impact

What this playbook delivers

Projected against the same workflow run manually, end to end

56%
Faster than the manual baseline
~7.4hrs
Saved per cycle
3
Source-of-truth artifacts

Estimated impact based on 20 candidates evaluated against 5 must-have criteria · Real savings vary with team volume and current process maturity

The challenge

Why the workflow breaks

Three patterns show up across teams using GenAI without a shared workflow

Evidence sprawl

Candidate evidence sits across resumes, notes, and screening artifacts, making it easy for important signals to disappear.

Criteria drift

Reviewers can compare candidates against different assumptions when must-have criteria are not applied consistently.

Ranking pressure

Shortlists often collapse into a ranked list before gaps, risks, and evidence quality have been reviewed.

Where GenAI helps

From confused intake to source-of-truth artifacts

Here’s what changes when the team uses this playbook

  • Reviewers skim applications independently

    Evidence Extraction Tables anchor the review

  • Candidate summaries emphasize different signals

    Candidate Summary Cards use the same structure

  • Risks are discussed late in the process

    Risk & Gaps Log records concerns before ranking

  • The shortlist is hard to explain

    Shortlist Table links buckets to evidence

How the playbook works

3 phases, one source of truth

Each phase produces an artifact the next phase builds on

1

Extract evidence

AI pulls candidate claims into an Evidence Extraction Table so reviewers can see what is supported by the source materials.

2

Verify gaps

The workflow produces Candidate Summary Cards and a Risk & Gaps Log before any shortlist decision is finalized.

3

Build shortlist

Verified evidence flows into the Shortlist Table (Buckets/Ranked), preserving the rationale behind each placement.

What you’ll produce

Sample artifacts from the workflow

Interim and final deliverables you can review and download

Candidate Summary Cards

One-page evidence profile per candidate, capturing role fit, strengths, and red flags. Every claim is traced to a source in the resume or screening note.

AiOS · HR03 · Step 2
Download Sample PDF

Risk & Gaps Log

Consolidated view of every must-have criterion with Partial or No Evidence. Grouped by criterion and by candidate to surface pool-level hiring risks before shortlist decisions.

AiOS · HR03 · Step 3
Download Sample PDF

Shortlist Table

Final ranked decision matrix assigning each candidate to Advance, Hold, or Decline. Every bucket placement is traceable to evidence counts and reflects hiring manager review.

AiOS · HR03 · Step 5
Download Sample PDF

Built into the workflow

Quality and risk checks at every step

Verification, data handling, and definition-of-done rules are part of the playbook — not afterthoughts

Quality

Enforces identical 5-criterion evaluation with verbatim evidence anchoring for every candidate, preventing the scoring drift that degrades manual screening after 10–15 resumes.

Risk handling

Prevents candidate PII from reaching unapproved AI tools and catches fabricated qualifications through a mandatory human cross-reference gate before the shortlist reaches the hiring manager.

LIVE ONLINE LABS

Join a live Lab for this workflow

An AGASI expert runs your team through the Screening & Candidate Shortlisting workflow on your own real work — so you leave with the method, not a recording.

  • Built around this exact playbook

    You internalise the method, not a recording.

  • Facilitated by an AGASI expert

    Live online coaching on prompts, verification, and where the workflow tends to break.

  • Public or private formats

    Join an open public Lab, or bring it in-house for your team.

I'm interested in

Get started

Run the Screening & Candidate Shortlisting playbook

Step-by-step prompts, role guidance, data-handling notes, and definition-of-done checks for every step of the workflow