The SignalHR
Comparison Hub · Applicant match scores

Applicant match scores: can you explain the number?

16 platforms that score or rank applicants, from the match score inside your ATS or HCM suite to talent-intelligence and sourcing tools. Rated on what you'd need to defend a score: explanations, recruiter control, human review, an off switch, published bias audits and AI certification. Start with the rules where you hire. No vendor paid to be here, rank higher or be described favorably.

Reviewed October 202616 platformsRules in 21 countries and the EUIndependent · unsponsoredAI-assisted
Step 1

Before you switch a score on: the rules where you hire

A match score can arrive in a product release without a separate purchase or review, and many hiring-AI rules are written with scores like these in mind. Pick the countries where you hire. Where a state, province or EU country adds its own rules, it opens underneath. Then say how the score will be used. What applies, the capabilities these rules point to and the questions for your vendor build below.

Americas
Europe, Middle East & Africa
Asia-Pacific
How the score will be used

What applies

Pick at least one place you hire.

Orientation, not legal advice. These are the 38 places we track in 21 countries and the EU, reviewed October 2026. Coverage thresholds, exemptions and pending rules vary, and anti-discrimination and privacy law apply everywhere, listed or not. Check with counsel where you hire.

Need this kept current? The Hiring-AI Governance Kit, in development, adds a monthly rules update and notice drafts. Tell us if you'd use it.

Step 2

Find your shortlist

Company size
Your ATS
Who you'll score (any that apply)
Must-haves (any that apply)

Step 3

Compare side by side

Tick Compare on two or three platforms in your shortlist, or pick them here:

The landscape

Four places a score comes from

Scores are now common in recruiting platforms, and they can be switched on in a release rather than bought. Ownership and litigation shape the risk: Workday owns HiredScore, SAP owns SmartRecruiters and is moving its recruiting there, Employ owns Lever, and Eightfold faces a class action over how candidate profiles are built. Assessments, video interviews and chatbot screening are out of scope here; they carry their own rules.

Score inside an ATS

The match or ranking feature in your applicant tracking system. Often the cheapest to switch on, and the one most often switched on without a governance review.

Greenhouse Talent Matching · Ashby criteria evaluation · Lever Talent Fit · iCIMS candidate ranking · SmartRecruiters Winston Match

Score inside an HCM suite

Scoring inside Workday, Oracle or SAP SuccessFactors. One data model with the HCM; the AI is frequently a separate SKU or consumption-priced.

Workday HiredScore · Oracle Recruiting recommendations · SAP SuccessFactors applicant screening

Talent intelligence platform

Skills-inference platforms that score applicants, past candidates and employees against roles, usually on top of your ATS.

Eightfold match score · Phenom Fit Score · Beamery Talent Match · Avature matching

Sourcing and screening tool

Tools that rank sourced prospects and, increasingly, inbound applicants, next to your ATS. Watch where the profile data comes from.

Gem AI Application Review · hireEZ Applicant Match · SeekOut AI Recommendations · Findem applicant matching

Deep dives

Platform profiles

Expand all

Each profile describes what a recruiter sees, who sets the criteria, what can happen automatically, and the public evidence behind the governance ratings. "Not documented" means we couldn't find it published, not that it's missing: ask.

Greenhouse Talent MatchingScore inside an ATSPrivate (Greenhouse Software)

Typical use: Teams on Greenhouse that want a governed, explainable first pass over inbound applicants.

Under 500 · 500–5,000 · 5,000+ employees · Scores inbound applicants, past candidates · Requires Greenhouse Recruiting

What the recruiter sees

Sorts applications into tiers of fit and shows which job criteria each résumé matches.

How the score is set

Recruiters choose which job criteria to prioritize and how much weight each gets.

Automatic actions

Cannot automatically advance or reject candidates.

Governance

  • Explains each scoreDocumented
  • Recruiter-set criteriaDocumented
  • No auto-rejectDocumented
  • Can be switched offNot documented
  • Published bias auditRecent, public
  • AI certificationCovers the product

Audit and certification evidence

ISO/IEC 42001 (February 2026). Warden AI continuously tests Talent Matching for disparate impact across 10 protected classes, gates each release on a passing audit and publishes results.

Strengths

  • Recruiter-weighted criteria with per-candidate matches shown
  • Warden AI tests every new version before release and publishes results on a public dashboard
  • ISO/IEC 42001 certified (February 2026)

Watch-outs

  • Whether and where it can be switched off isn't documented publicly
  • Past-candidate matching arrived in June 2026; check which plan includes it
  • Tiers are not a decision: make sure recruiters still read below the top tier

How it's priced

Part of Greenhouse's recruiting plans; confirm which tier includes Talent Matching.

Ask in the demo

  • Show the criteria and weights behind one candidate's tier, and how a recruiter changes them.
  • Can we turn Talent Matching off for one country or job, and is that change logged?

Verified October 2026 · Sources: Greenhouse bias audit statement · Greenhouse release notes, June 2026 · Greenhouse ISO/IEC 42001

Ashby criteria evaluationScore inside an ATSPrivate (Ashby)

Typical use: Teams that want AI to check criteria without collapsing candidates into a single number.

Under 500 · 500–5,000 employees · Scores inbound applicants, past candidates · Requires Ashby

What the recruiter sees

Meets, does not meet or uncertain for each recruiter-defined criterion. No overall score or ranking.

How the score is set

Recruiters define criteria per role; a language model checks each résumé against each criterion.

Automatic actions

Outputs are reviewed criterion by criterion; automatic actions aren't documented.

Governance

  • Explains each scoreDocumented
  • Recruiter-set criteriaDocumented
  • No auto-rejectPartial
  • Can be switched offPartial
  • Published bias auditOlder or narrower
  • AI certificationNone found

Audit and certification evidence

FairNow audited the criteria-evaluation model under NYC Local Law 144 (August 2024): no group below the four-fifths threshold. No AI certification found.

Strengths

  • No single score to anchor on: recruiters see criterion-level results
  • 'Uncertain' when the résumé doesn't say, instead of a guess
  • Independent NYC bias audit published (August 2024)

Watch-outs

  • The only published audit is from August 2024 and covers one model
  • AI Talent Rediscovery (May 2026) uses metered credits
  • AI tools are opt-in; customers carry notice and consent duties

How it's priced

AI features in Ashby's plans with credit allowances by tier; rediscovery uses credits per candidate returned.

Ask in the demo

  • When will you publish a 2025 or 2026 audit, and will it cover rediscovery?
  • Can a 'does not meet' result trigger any automatic action in our workflows?

Verified October 2026 · Sources: Ashby bias audit by FairNow, August 2024 · Ashby terms for AI features · Ashby: AI Talent Rediscovery

Lever Talent FitScore inside an ATSEmploy Inc.

Typical use: Lever customers that want a quick shortlist from inbound volume.

Under 500 · 500–5,000 employees · Scores inbound applicants · Requires Lever

What the recruiter sees

A 'Top Candidates' view with labels such as 'AI pick' and 'Best fit' and a count of matched criteria.

How the score is set

AI matching against customizable criteria.

Automatic actions

Lever says recruiters always have the final say; automatic actions aren't documented.

Governance

  • Explains each scorePartial
  • Recruiter-set criteriaDocumented
  • No auto-rejectPartial
  • Can be switched offNot documented
  • Published bias auditNone found
  • AI certificationNone found

Audit and certification evidence

Employ cites IBM watsonx governance and independent bias audits, with no named auditor or published results. No AI certification found.

Strengths

  • Customizable matching criteria
  • Built into the core product
  • AI Screening Companion adds candidate-loss alerts and follow-ups

Watch-outs

  • Says recruiters can see why candidates are recommended, but the reasons aren't documented
  • Employ claims independent bias audits but publishes none
  • Screening add-ons (VONQ) bring a second vendor's model into the process

How it's priced

Talent Fit is part of Lever's built-in AI; AI screening by VONQ is an add-on. Confirm entitlements in the order form.

Ask in the demo

  • Show the reasons behind one 'AI pick' and who set the criteria.
  • Send the latest independent audit of Talent Fit, with the auditor's name.

Verified October 2026 · Sources: Lever AI features · Employ: AI Screening Companion

iCIMS candidate rankingScore inside an ATSPrivate (iCIMS)

Typical use: Enterprise and high-volume teams on iCIMS that want ranking and sourcing on the same record.

500–5,000 · 5,000+ employees · Scores inbound applicants, past candidates, sourced prospects · Requires iCIMS

What the recruiter sees

Ranked candidates and role matches based on skills and experience.

How the score is set

Ranks candidates on skills and experience; the CXM AI Sourcing Agent searches your iCIMS database.

Automatic actions

Recruiters set how much autonomy agents get; automatic rejection isn't documented.

Governance

  • Explains each scoreNot documented
  • Recruiter-set criteriaPartial
  • No auto-rejectPartial
  • Can be switched offNot documented
  • Published bias auditOlder or narrower
  • AI certificationCovers the product

Audit and certification evidence

TrustArc TRUSTe Responsible AI certification (March 2025) and an independent NYC Local Law 144 audit by BABL AI (December 2023).

Strengths

  • TrustArc Responsible AI certification (March 2025)
  • Agent autonomy is configurable by the customer
  • Ranking spans applicants, past candidates and the talent database

Watch-outs

  • How a ranking is explained isn't documented publicly
  • The published bias audit dates from December 2023
  • The AI Sourcing Agent needs the CXM package

How it's priced

Ranking in the iCIMS platform; the AI Sourcing Agent is included in CXM, a separately licensed package.

Ask in the demo

  • Show why one candidate ranks above another, in the recruiter's view.
  • When was the last bias audit, and does it cover the 2026 ranking and agent features?

Verified October 2026 · Sources: iCIMS AI recruiting search and automation · BABL AI on its iCIMS audit · iCIMS Spring Release 2026

SmartRecruiters Winston MatchScore inside an ATSSAP (since September 2025)

Typical use: SmartRecruiters customers, on SAP or not, that want an explained score across a large applicant flow.

500–5,000 · 5,000+ employees · Scores inbound applicants · Requires the SmartRecruiters ATS

What the recruiter sees

A match score labeled Very high, High, Medium or Low, with a plain-language explanation.

How the score is set

Separate models score skills, experience, education, career progression and title relevance; a weighted total combines them, with weights based on hiring outcomes.

Automatic actions

Automatic rejection isn't documented.

Governance

  • Explains each scoreDocumented
  • Recruiter-set criteriaNot documented
  • No auto-rejectPartial
  • Can be switched offNot documented
  • Published bias auditNone found
  • AI certificationCovers the product

Audit and certification evidence

Announced ISO/IEC 42001 in August 2025 (certifier not named) and says its AI passes an annual NYC Local Law 144 audit; no report found.

Strengths

  • Explanations of why a candidate does or doesn't match
  • Multi-dimension scoring with a 35,000+ skill taxonomy, per SmartRecruiters
  • ISO/IEC 42001 announced (August 2025)

Watch-outs

  • Weights come from hiring outcomes, so ask what history trained them
  • Claims an annual NYC audit, but no report is published
  • A 'Low' label can work as a de facto rejection; decide how recruiters treat it

How it's priced

Part of Winston AI, sold by tier; confirm Winston Match is in your package.

Ask in the demo

  • Which hiring outcomes set the weights, and can we see the weights for one of our jobs?
  • Publish the latest bias audit of Winston Match, with the auditor's name.

Verified October 2026 · Sources: SmartRecruiters Winston Match · SmartRecruiters ISO 42001 (community post)

Workday HiredScoreScore inside an HCM suiteWorkday

Typical use: Workday customers grading high volumes of applicants and rediscovering past ones inside the requisition.

500–5,000 · 5,000+ employees · Scores inbound applicants, past candidates, employees · Requires Workday Recruiting

What the recruiter sees

Grades from A to D, with A the closest match to the job's requirements.

How the score is set

Compares candidates' reported qualifications with the job's requirements; Fetch resurfaces past applicants and CRM leads.

Automatic actions

Outcomes depend on your workflow design and human review; automatic rejection isn't documented.

Governance

  • Explains each scoreNot documented
  • Recruiter-set criteriaNot documented
  • No auto-rejectPartial
  • Can be switched offNot documented
  • Published bias auditRecent, public
  • AI certificationCovers the product

Audit and certification evidence

Secretariat analyzed Spotlight grades on Workday's own applicants (testing completed March 20, 2026): no evidence of disparate impact, published for transparency only. ISO/IEC 42001 and NIST AI RMF attestation (June 2025).

Strengths

  • Published independent bias analysis of Spotlight grades (March 2026)
  • ISO/IEC 42001 and NIST AI RMF attestation
  • Rediscovery of past applicants and CRM leads inside the requisition

Watch-outs

  • The bias analysis covers Workday's own hiring data, not yours; customers still run their own NYC audits
  • HiredScore is a separate SKU, and agents consume Flex Credits
  • How a grade is explained to recruiters isn't documented publicly

How it's priced

HiredScore AI for Recruiting is a separate SKU; agents use usage-based Flex Credits.

Ask in the demo

  • Show what a recruiter sees next to a B grade, and what would make it an A.
  • Will you support our own NYC audit with grade and outcome data for our jobs?

Verified October 2026 · Sources: Workday: bias mitigation for HiredScore Spotlight · HiredScore Embedded Fetch · Workday AI certifications

Oracle Recruiting recommendationsScore inside an HCM suiteOracle

Typical use: Oracle Cloud HCM customers that want recommendations across external and internal talent.

5,000+ employees · Scores inbound applicants, sourced prospects, employees · Requires Oracle Recruiting

What the recruiter sees

AI recommendations of best-fit prospects and applicants, and skill-based recommendations for internal and external candidates.

How the score is set

Skills-based matching inside Oracle Recruiting.

Automatic actions

Automation can move candidates through business processes; how recommendations feed those steps isn't documented.

Governance

  • Explains each scoreNot documented
  • Recruiter-set criteriaNot documented
  • No auto-rejectNot documented
  • Can be switched offNot documented
  • Published bias auditNone found
  • AI certificationCovers the product

Audit and certification evidence

Oracle's ISO/IEC 42001:2023 certificate from Schellman covers Fusion Cloud Applications (June 2026). No recruiting-specific bias audit found.

Strengths

  • ISO/IEC 42001 certificate covering Fusion Cloud Applications
  • Recommendations span internal and external candidates
  • Generative-AI candidate search added in the 26B update (May 2026)

Watch-outs

  • Explanations, recruiter controls and bias testing aren't documented publicly
  • Some AI and engagement features sit in Recruiting Booster, an add-on
  • Oracle AI World (October 26–29, 2026) may change the roadmap

How it's priced

Included in Oracle Recruiting; Recruiting Booster is a separately priced add-on.

Ask in the demo

  • Show how a recommendation is explained to a recruiter, and how it's tested for adverse impact.
  • Which recommendation features need Booster or extra AI pricing?

Verified October 2026 · Sources: Oracle Recruiting and Recruiting Booster · Oracle: raising the bar for trustworthy AI

SAP SuccessFactors applicant screeningScore inside an HCM suiteSAP

Typical use: SuccessFactors customers that want skills-based screening inside the suite.

500–5,000 · 5,000+ employees · Scores inbound applicants · Requires SAP SuccessFactors Recruiting

What the recruiter sees

A skills match between the applicant and the requisition, including related skills the applicant may have or could acquire.

How the score is set

Extracts skills from applicant data and compares them with the requisition's skills, which can be augmented.

Automatic actions

Not documented.

Governance

  • Explains each scoreNot documented
  • Recruiter-set criteriaPartial
  • No auto-rejectNot documented
  • Can be switched offNot documented
  • Published bias auditNone found
  • AI certificationNarrower scope

Audit and certification evidence

SAP holds ISO/IEC 42001 for its AI governance, Joule and AI platform, but SuccessFactors isn't named in scope. No bias audit found.

Strengths

  • Skills-based rather than keyword matching
  • Works with SuccessFactors Recruiting
  • Covered by SAP's AI ethics review process

Watch-outs

  • Sold only in Joule Premium for HCM, paid in AI units
  • SAP says it isn't available in every country or region
  • SAP's recruiting investment is moving to SmartRecruiters

How it's priced

Joule Premium for HCM, purchased with AI units.

Ask in the demo

  • Is applicant screening available in all the countries we hire in?
  • Show the skills match for one applicant and how a recruiter overrides it.

Verified October 2026 · Sources: SAP SuccessFactors applicant screening · SAP responsible AI

Eightfold match scoreTalent intelligence platformPrivate (Eightfold AI)

Typical use: Enterprises scoring external applicants, past candidates and employees on one skills model.

5,000+ employees · Scores inbound applicants, past candidates, sourced prospects, employees · Documented ATS integrations: SAP SuccessFactors, Greenhouse

What the recruiter sees

A score from 0 to 5 in half steps for each candidate and job pair, with a breakdown by skill, work and title relevance.

How the score is set

A machine-learning model infers skills and compares candidates with job requirements; scores change as either side's information changes.

Automatic actions

Eightfold says the model doesn't replace human decisions; recruiters can hire anyone and candidates can apply anywhere.

Governance

  • Explains each scoreDocumented
  • Recruiter-set criteriaPartial
  • No auto-rejectDocumented
  • Can be switched offNot documented
  • Published bias auditRecent, public
  • AI certificationCovers the product

Audit and certification evidence

BABL AI audited the Eightfold Matching Model under NYC Local Law 144 (March 6, 2026), with a public summary. ISO/IEC 42001, SOC 2 Type II and FedRAMP Moderate.

Strengths

  • Hover breakdowns of skill, work and title relevance
  • Independent NYC audit of the matching model (March 2026)
  • ISO/IEC 42001, SOC 2 and FedRAMP Moderate

Watch-outs

  • Faces an FCRA class action alleging its profiles and scores are consumer reports built partly from scraped data (filed January 2026)
  • Ask what outside data enters your candidates' profiles, and whether you can switch it off
  • A 0–5 score invites cut-offs; decide how recruiters may use it

How it's priced

Quote-based enterprise subscription, typically priced by employee count.

Ask in the demo

  • Which outside data sources feed our candidates' profiles, and can we turn each off?
  • Will you share score and outcome data for our jobs so we can run our own audit?

Verified October 2026 · Sources: Eightfold matching model bias audit · Epstein Becker Green: the Eightfold litigation

Phenom Fit ScoreTalent intelligence platformPrivate (Phenom)

Typical use: Enterprises on Workday or SuccessFactors that want explained scoring across CRM, career site and internal mobility.

500–5,000 · 5,000+ employees · Scores inbound applicants, past candidates, sourced prospects, employees · Documented ATS integrations: Workday, SAP SuccessFactors

What the recruiter sees

Ranks candidates on skills, title and location, with a reason behind every grade.

How the score is set

Recruiters describe requirements in plain language; a skills ontology matches related skills, titles and roles.

Automatic actions

Recruiters can override; separately, workflows can schedule interviews when a candidate qualifies and turn segments into scored shortlists.

Governance

  • Explains each scoreDocumented
  • Recruiter-set criteriaPartial
  • No auto-rejectPartial
  • Can be switched offNot documented
  • Published bias auditNone found
  • AI certificationNone found

Audit and certification evidence

Announced an EU AI Act conformity assessment and NIST AI RMF alignment in March 2026, both self-assessed. No ISO/IEC 42001 or published bias audit found.

Strengths

  • A reason shown for every grade
  • Recruiter override
  • Ranks applicants, CRM leads and employees together

Watch-outs

  • Automations can act on scores (scheduling, scored shortlists); map where a human reviews
  • No published bias audit; governance evidence is a self-assessed EU AI Act conformity assessment
  • Plum assessments (acquired April 2026) may feed scoring; ask how

How it's priced

Quote-based subscription, reported as per employee per month; varies by modules.

Ask in the demo

  • Which automations can act on a Fit Score without a recruiter, and can we turn them off?
  • Publish independent adverse-impact results for Fit Score on hiring data.

Verified October 2026 · Sources: Phenom Talent CRM · Phenom: AI that redesigns work operations, March 2026

Beamery Talent MatchTalent intelligence platformPrivate (Beamery)

Typical use: Workday and SuccessFactors enterprises that want ranked applicants and rediscovery with strong audit evidence.

5,000+ employees · Scores inbound applicants, past candidates, sourced prospects · Documented ATS integrations: Workday, SAP SuccessFactors, Greenhouse, SmartRecruiters

What the recruiter sees

A match score that ranks applicants in Applicant Review and suggests contacts from your talent pool.

How the score is set

Matching against calibrated vacancies, with recruiter-chosen filters such as skills and years of experience.

Automatic actions

Filters are never applied automatically; rejection is a one-click recruiter action.

Governance

  • Explains each scoreNot documented
  • Recruiter-set criteriaPartial
  • No auto-rejectDocumented
  • Can be switched offDocumented
  • Published bias auditRecent, public
  • AI certificationCovers the product

Audit and certification evidence

Warden AI has independently audited Beamery's AI for bias by sex and ethnicity since August 2024, with a live dashboard. ISO/IEC 42001 announced August 2025 (certifier and scope not named).

Strengths

  • Continuous independent bias monitoring by Warden AI since August 2024
  • Match scores can be hidden at account level
  • ISO/IEC 42001 (August 2025)

Watch-outs

  • How a score is explained isn't documented publicly
  • AI matching requires a separate AI addendum
  • Leadership changes and feature retirements during 2025–26

How it's priced

Quote-based modular suites; AI matching sits behind an AI addendum.

Ask in the demo

  • Show the reasons behind one applicant's rank.
  • Does the ISO 42001 scope and the Warden audit cover Talent Match in Applicant Review?

Verified October 2026 · Sources: Beamery release note, September 2026 · Beamery: Warden AI audit results · Beamery ISO 42001

Avature matchingTalent intelligence platformPrivate (Avature)

Typical use: Global enterprises on Avature that want configurable matching they can switch on by use case.

5,000+ employees · Scores inbound applicants, past candidates, employees · No ATS integrations documented by name (says it works with others)

What the recruiter sees

AI candidate suggestions; Avature describes the AI as fully explainable and configurable.

How the score is set

Large language models plus Avature's fine-tuned, industry-specific models.

Automatic actions

Not documented.

Governance

  • Explains each scorePartial
  • Recruiter-set criteriaPartial
  • No auto-rejectNot documented
  • Can be switched offDocumented
  • Published bias auditNone found
  • AI certificationNone found

Audit and certification evidence

Says it follows the NIST AI RMF, runs annual independent audits including a New York City AEDT audit, and aligns with the EU AI Act. No published audit summary or certification found.

Strengths

  • AI can be enabled and disabled
  • Described as explainable and adjustable by end users
  • Matching across applicants, past candidates and employees

Watch-outs

  • Claims annual independent audits, including a New York City one, without a published summary
  • How explanations appear isn't documented publicly
  • No ATS integrations documented by name

How it's priced

Quote-based; solutions licensed separately.

Ask in the demo

  • Send the latest bias audit summary and the auditor's name.
  • Show an explanation for one suggestion, as a recruiter sees it.

Verified October 2026 · Sources: Avature AI

Gem AI Application ReviewSourcing and screening toolPrivate (Gem)

Typical use: Recruiting teams that want transparent, recruiter-controlled ranking on top of their existing ATS.

Under 500 · 500–5,000 · 5,000+ employees · Scores inbound applicants, sourced prospects · Documented ATS integrations: Workday, SAP SuccessFactors, iCIMS, Greenhouse, Lever, SmartRecruiters, Ashby

What the recruiter sees

A percentage match for each applicant, with a score and explanation for every criterion; the total averages them.

How the score is set

Recruiters define the criteria and can change them and re-score at any time; personal data is removed before scoring and subjective criteria trigger bias warnings.

Automatic actions

Doesn't make advancement or rejection decisions.

Governance

  • Explains each scoreDocumented
  • Recruiter-set criteriaDocumented
  • No auto-rejectDocumented
  • Can be switched offPartial
  • Published bias auditOlder or narrower
  • AI certificationNone found

Audit and certification evidence

BABL AI audited Gem's AI ranking under NYC Local Law 144 (November 2024); lowest impact ratio 0.834, with the summary hosted by a customer. No AI certification found.

Strengths

  • Per-criterion scores and explanations
  • Bias warnings on subjective criteria; personal data removed before scoring
  • Works across seven major ATS platforms

Watch-outs

  • The only published audit (November 2024) is hosted on a customer's site
  • Turned on per job; no account-wide switch documented
  • Paid add-on; AI features need AI-Powered seats

How it's priced

Paid add-on through your account manager; AI features need AI-Powered seats.

Ask in the demo

  • Where is the 2025 or 2026 audit, and does it cover today's Application Review?
  • Can we disable Application Review for specific countries or states?

Verified October 2026 · Sources: Gem: AI features overview · Gem bias audit summary by BABL AI (hosted by Okta) · Gem ATS integrations

hireEZ Applicant MatchSourcing and screening toolPrivate (hireEZ, formerly Hiretual)

Typical use: Sourcing-led teams that want outbound search and inbound shortlisting from one vendor, with published audit reports.

500–5,000 · 5,000+ employees · Scores inbound applicants, past candidates, sourced prospects · Documented ATS integrations: Lever, iCIMS

What the recruiter sees

AI matching that shortlists the best-fit applicants, alongside sourcing across 45+ platforms.

How the score is set

Matching across sourced profiles and inbound applicants; criteria details aren't documented publicly.

Automatic actions

hireEZ says its tools don't make employment decisions; those are always made by users.

Governance

  • Explains each scoreNot documented
  • Recruiter-set criteriaNot documented
  • No auto-rejectPartial
  • Can be switched offNot documented
  • Published bias auditOlder or narrower
  • AI certificationNone found

Audit and certification evidence

Independent bias assessments by Armilla AI and Warden AI, with a summary and three audit reports published (sourcing, Applicant Match and AI screening); dates not stated. No AI certification found.

Strengths

  • Published independent bias assessments by Armilla AI and Warden AI covering sourcing, Applicant Match and AI screening
  • Sourcing across 45+ platforms
  • A sourcing partner for Workday's Talent Acquisition agent (early adopters)

Watch-outs

  • Audit reports carry no date on the page we reviewed
  • How matches are explained and who sets criteria aren't documented
  • Ask where external profile data comes from

How it's priced

Not published on the pages we reviewed; quote-based.

Ask in the demo

  • Which data sources feed sourced profiles, and how are they licensed?
  • Show the criteria behind one shortlisted applicant and how a recruiter changes them.

Verified October 2026 · Sources: hireEZ AI values for recruiting · hireEZ–Lever integration · hireEZ–iCIMS integration

SeekOut AI RecommendationsSourcing and screening toolPrivate (SeekOut)

Typical use: Talent teams sourcing hard-to-fill roles who want scorecard-based evaluation of large candidate pools.

500–5,000 · 5,000+ employees · Scores sourced prospects, past candidates · Documented ATS integrations: Greenhouse

What the recruiter sees

Scorecard evaluations per criterion, filterable by score range, surfaced as AI Recommendations.

How the score is set

Matches people to a job description and evaluates them against scorecard criteria; SeekOut Spot adds expert recruiters who validate AI matches.

Automatic actions

Administrators can turn off automatic AI evaluation and recommendations (June 2026).

Governance

  • Explains each scoreDocumented
  • Recruiter-set criteriaPartial
  • No auto-rejectPartial
  • Can be switched offDocumented
  • Published bias auditNone found
  • AI certificationNone found

Audit and certification evidence

No published bias audit or AI certification found as of October 2026.

Strengths

  • Criterion-level scorecards with descriptions
  • Admins can switch automatic evaluation off while keeping search
  • Bulk evaluation of up to 500 top matches

Watch-outs

  • No bias audit or responsible-AI note in 2025–26 release notes
  • Score-range filters can become de facto cut-offs
  • Check how candidates in your ATS are evaluated versus sourced profiles

How it's priced

Not published on the pages we reviewed; quote-based.

Ask in the demo

  • Publish an independent adverse-impact audit of AI Recommendations.
  • Who writes the scorecard criteria, and can we lock them per job?

Verified October 2026 · Sources: SeekOut release notes

Findem applicant matchingSourcing and screening toolPrivate (Findem)

Typical use: Teams that want sourcing and applicant matching with a recent, public independent audit.

500–5,000 · 5,000+ employees · Scores inbound applicants, sourced prospects · Documented ATS integrations: Workday, SAP SuccessFactors, iCIMS, Greenhouse, Lever, Ashby, Jobvite

What the recruiter sees

Evaluates each profile against employer-defined requirements, such as skills and experience.

How the score is set

Each profile is assessed individually against the same employer-defined criteria.

Automatic actions

Findem says hiring decisions remain human-led.

Governance

  • Explains each scoreNot documented
  • Recruiter-set criteriaDocumented
  • No auto-rejectPartial
  • Can be switched offNot documented
  • Published bias auditRecent, public
  • AI certificationNone found

Audit and certification evidence

Warden AI audited Findem's applicant-matching practices (completed December 2025, announced January 2026): met Warden's fairness thresholds, with results on a public dashboard. No AI certification found.

Strengths

  • Independent Warden AI audit (December 2025) with a live public dashboard
  • Employer-defined criteria applied the same way to every profile
  • Integrates with Workday, SuccessFactors, iCIMS, Greenhouse, Lever and Ashby

Watch-outs

  • How results are explained isn't documented publicly
  • Ask where enriched profile data comes from
  • A sourcing partner for Workday's agent; check which features you're actually buying

How it's priced

Not published on the pages we reviewed; quote-based.

Ask in the demo

  • Show the explanation a recruiter sees for one match.
  • Which profile data comes from outside sources, and can candidates see or correct it?

Verified October 2026 · Sources: Findem: AI bias audit with Warden AI · Findem integrations

Free to use

The 12-question match-score RFP bank

Questions that show whether you could explain a score to a candidate, a regulator or a works council. A vendor's answers show how much of a score you'd be able to explain yourself.

  1. Show, for one of our jobs, exactly what a recruiter sees next to a candidate's score, and what would change it.
  2. Who sets the criteria and weights: us, the model, or past hiring outcomes? Can we see and lock them per job?
  3. Can the score automatically reject, advance, hide or contact anyone? What are the defaults, and who can change them?
  4. Can we switch scoring off by country, state or job, and is every change logged?
  5. Send the latest independent bias audit: auditor, date, data, the features covered and the impact ratios.
  6. Will you give us score and outcome data for our own jobs so we can run our own audit?
  7. Which data feeds a candidate's profile beyond what they submitted, and can we switch each source off?
  8. What do you provide for our candidate notices (NYC, Illinois, Colorado, California, Ontario, EU)?
  9. Which AI certification do you hold, who issued it, and does its scope name this feature?
  10. How do you test a new model version before release, and how will we know it changed?
  11. Is scoring included in our license, a separate SKU, or consumption-priced?
  12. If a candidate asks why they were rejected, what can we tell them from your system?
Independence

How we review

  • No vendor pays. No sponsorships, affiliate links or paid placement, and no vendor reviews its profile before publication.
  • No recommendations. We describe what each platform documents. We don't rank platforms or tell you which to choose: the finder only shows which platforms match the filters you pick.
  • Scope. Scores, grades and rankings computed from applications, résumés and profiles. Assessments, games, video interviews and conversational screening are out of scope: they work differently and carry their own rules.
  • Ratings reflect public documentation. "Documented" means the vendor documents it publicly; "Partial" means claimed without detail, or limited; "Not documented" means we found nothing published, so ask. We rate the score as sold, not a test of the product.
  • Bias audits. "Recent, public" means an independent audit or continuous monitoring of the scoring feature, published within about the last 18 months. "Older or narrower" means older, undated, a single model, or hosted by a customer. Audit claims without a published summary count as none found.
  • AI certification. "Covers the product" means an independent certification of the vendor's AI management, such as ISO/IEC 42001 or TrustArc, that covers the vendor or the product. "Narrower scope" means its scope doesn't name the product.
  • Rules where you hire. 38 jurisdictions in 21 countries and the EU, each linked to a source. Picking a state, province or EU country also shows the national or EU-wide rules. Where we don't track a place individually, the check says so instead of going quiet. It's orientation, not legal advice.
  • ATS integrations. Listed only when we found them on a marketplace, partner or help page, not from a vendor's "integrates with every ATS" claim.
  • Pricing. We describe how each vendor charges and where costs grow, not list prices, because enterprise prices are negotiated and change.
  • Freshness. Every profile shows when it was verified. We re-verify quarterly and after major vendor or legal news. Spot something out of date? Tell us and we'll fix it with a note. Every material change is in the change log.

Get the signal each week.

The weekly signal: what changed in HR tech, what it may mean, and what to check. Free forever. See the buyer kits we're building.

No spam. Unsubscribe anytime.