HeyMilo
HireAIScore rates HeyMilo 61 out of 100 (grade D, Concerning) under rubric v1.0, last reviewed June 7, 2026. HeyMilo ranks 7 of 160 vendors rated in Screening AI, against a category average of 46. That score is drawn from 21 evidence items across 7 rubric criteria for HeyMilo — 18 with a cited source, 3 recording that nothing was located.
Score breakdown
HeyMilo, criterion by criterion
Each rubric criterion is scored out of 100 on the cited evidence, then weighted into the total.
Weighted contribution shown to the right of each bar.
Article 11 Technical Documentation
Weight 20% · category avg 49
58
+11.6
Bias Audit Transparency
Weight 18% · category avg 41
83
+14.9
FRIA Support
Weight 15% · category avg 31
42
+6.3
Data Governance Disclosure
Weight 15% · category avg 53
62
+9.3
Human Oversight Design
Weight 12% · category avg 56
64
+7.7
Post-Market Monitoring
Weight 12% · category avg 40
46
+5.5
Customer Documentation
Weight 8% · category avg 58
66
+5.3
Total weighted score
60.6(rounded to 61 for display)
Category avg is the mean raw score on that criterion across the 160 Screening AI vendors in scope of this rubric, this one included.
Strongest criterion
Bias Audit Transparency
83raw / 100
Contributes 14.9 to the total. 83 against a 41 category average.
Weakest criterion
FRIA Support
42raw / 100
Contributes 6.3 to the total. 42 against a 31 category average.
Cited evidence
The evidence behind the score
Every score is backed by at least one cited piece of evidence. Open a criterion to see what was found, when, and where.
Download diligence recordEvidence ledger
- Items
- 21
- Documentation
- 11
- Audit report
- 3
- Public statement
- 3
- Product feature
- 1
- Absence
- 3
- With a source URL
- 18 of 21
- Source hostnames
- 2
- Fewest items
- 3
- Article 11 Technical Documentation
These figures measure how thoroughly HeyMilo was reviewed, not how HeyMilo performed — an absence row, recording that nothing was located, is counted like any other item.
Article 11 Technical Documentation
3 items58
- DocumentationCaptured Jun 7, 2026
Responsible AI page (by CEO and CTO, updated 2026-03-04) states the AI is advisory/decision-support and details per-question scores, AI-generated rationale, timestamped candidate quotes, and full transcripts as an explainability framework.
- Public statementCaptured Jun 7, 2026
Page claims EU AI Act classification analysis for standard configurations and Article 4 AI-literacy training for all employees, but no downloadable model/system card or ISO 42001 is published.
- AbsenceCaptured Jun 7, 2026
No public model card, system card, or Article 11-style technical documentation pack was found on heymilo.ai or its trust portals; the full technical measures are described only as available via the gated Trust Center.
Bias Audit Transparency
3 items83
- Audit reportCaptured Jun 7, 2026
Warden AI public assurance dashboard shows monthly bias audits (latest 28 May 2026, sample size 17,442) returning 'Clear' across 15 protected attributes including sex, race/ethnicity, and intersectional sex x race, with a downloadable report.
- Audit reportCaptured Jun 7, 2026
Dashboard maps bias-audit compliance to NYC Local Law 144, Colorado SB 205, the EU AI Act (Reg 2024/1689), and California FEHA, each with a linked report.
- Public statementCaptured Jun 7, 2026
Responsible AI page confirms the Warden AI partnership for independent monthly bias audits and states results are published publicly and continuously.
FRIA Support
3 items42
- DocumentationCaptured Jun 7, 2026
Blog explains EU AI Act deployer obligations: Article 14 human oversight, the provider-vs-deployer distinction, and Annex III Section 4 high-risk classification for recruitment.
https://www.heymilo.ai/blog/how-the-eu-ai-act-changes-recruitment-and-what-employers-need-to-know
- DocumentationCaptured Jun 7, 2026
Blog tracks the EU AI Act high-risk employment compliance timeline and deployer responsibilities, indicating active deployer-facing regulatory guidance.
https://www.heymilo.ai/blog/eu-ai-act-2027-hiring-deadline-extension
- AbsenceCaptured Jun 7, 2026
No Article 27 Fundamental Rights Impact Assessment template, FRIA questionnaire, or deployer FRIA guidance was found; the EU AI Act blogs cover Articles 14/26 but never mention FRIA or Article 27.
Data Governance Disclosure
3 items62
- DocumentationCaptured Jun 7, 2026
States customer/candidate data is never used to train or fine-tune models, no cross-customer learning, data minimization by design, PII masking/redaction before reaching models, and ~7-day prompt-log retention with restricted access.
- DocumentationCaptured Jun 7, 2026
Privacy policy confirms interview recordings, video, transcripts, and application materials are not used to train or fine-tune general-purpose AI models, and lists CCPA/CPRA data categories (biometric: No).
- DocumentationCaptured Jun 7, 2026
Public GDPR-compliant Data Processing Agreement names Redsnack Technologies Inc (dba HeyMilo AI) as controller and references SCCs, subprocessor controls, and EU data-protection obligations.
Human Oversight Design
3 items64
- DocumentationCaptured Jun 7, 2026
States human-in-the-loop is enforced at every stage: AI is advisory only, recruiters define rubrics and thresholds, every score/summary/recommendation is subject to human review, and the recruiter makes the final call.
- DocumentationCaptured Jun 7, 2026
Describes in-product explainability and auditability: per-question scores, AI rationale grounded in candidate statements, candidate quotes with timestamps, full transcripts, and explicit, timestamped, revocable candidate consent.
- DocumentationCaptured Jun 7, 2026
Blog frames a named human reviewer with override/approve/escalate authority and logged overrides as a design requirement under EU AI Act Article 14.
https://www.heymilo.ai/blog/how-the-eu-ai-act-changes-recruitment-and-what-employers-need-to-know
Post-Market Monitoring
3 items46
- Audit reportCaptured Jun 7, 2026
Warden AI dashboard explicitly provides 'post-market monitoring' with monthly audit cadence and continuously updated public results, functioning as an ongoing fairness-monitoring surface.
- Public statementCaptured Jun 7, 2026
States formal incident-response plans with defined severity levels and a 48-hour customer-notification commitment for confirmed incidents, plus a published contact (hello@heymilo.ai) for audit/governance questions.
- AbsenceCaptured Jun 7, 2026
No public security status page, vulnerability-disclosure/security.txt, or model-update changelog was found; SOC 2 reporting and detailed monitoring evidence sit behind the gated Delve trust portal.
Customer Documentation
3 items66
- DocumentationCaptured Jun 7, 2026
Responsible AI hub links to the AI Assurance Dashboard, Trust Center, SOC 2 and GDPR pages, Privacy Policy, Data Processing Agreement, and Subprocessors list.
- DocumentationCaptured Jun 7, 2026
Long-form buyer/deployer guide on fair hiring software referencing NYC LL144, EU AI Act, Colorado, Illinois, EEOC Title VII, and the four-fifths rule, aimed at procurement and compliance evaluation.
https://www.heymilo.ai/blog/how-we-ensure-our-applicant-screening-software-stays-fair
- Product featureCaptured Jun 7, 2026
Publicly accessible Data Processing Agreement and dedicated product, solution, and ATS-integration pages (Bullhorn, Greenhouse, Ashby, Workday, iCIMS, etc.) provide concrete deployer-facing documentation.
Editorial notes
Company overview
HeyMilo (legal entity Redsnack Technologies Inc) is a New York-based, seed-stage AI recruiting company founded in 2023 by CEO Sabashan Ragavan and CTO Ramie Raufdeen. Its product is an 'AI recruiter' for high-volume hiring teams (staffing agencies, BPOs, franchises, and enterprises) that sources, pre-screens, and conducts adaptive voice/video/phone interviews in 14+ languages, then generates per-question scores, written rationale, transcripts, and recordings that sync to ATS platforms such as Bullhorn, Greenhouse, Ashby, Workday, and iCIMS. The company positions itself explicitly around 'fair hiring software,' framing auditability, explainability, and human-in-the-loop review as core product architecture rather than add-ons.
Regulatory exposure
HeyMilo's AI scores and ranks candidates in employment screening, placing it squarely within the high-risk category of the EU AI Act (Annex III) and within scope of NYC Local Law 144, Colorado SB 205, Illinois, and California FEHA. The company is unusually transparent for its stage: an independent Warden AI assurance dashboard publishes continuous monthly bias audits (latest 28 May 2026, sample 17,442) mapped to NYC LL144, Colorado SB 205, the EU AI Act, and California FEHA, and a detailed Responsible AI page documents human-in-the-loop design, explainability, data-governance commitments, SOC 2 Type I/II, GDPR, and EU AI Act classification work. The main residual exposure is the absence of a Fundamental Rights Impact Assessment (Article 27) template or deployer FRIA tooling, and the fact that core compliance artifacts (SOC 2 report, technical pack) sit behind gated trust portals.
Path to a higher score
To raise its score, HeyMilo should publish a public model/system card or Article 11-style technical documentation pack and an ISO 42001 certification rather than relying on a marketing-style Responsible AI page plus gated trust portals. It should add explicit EU AI Act Article 27 Fundamental Rights Impact Assessment guidance or a deployer FRIA template (its existing Article 14/26 blogs stop short of this). Making the SOC 2 report and Warden AI methodology downloadable without an NDA, and adding a public security.txt / status page / model-update changelog for post-market monitoring, would convert several strong-but-gated signals into fully verifiable ones.
Regulatory frame
What applies to screening AI.
Treated as high-risk under Annex III §4 when the screening output is decisive. EU AI Act Article 13 transparency obligations apply; deployers must give candidates a way to contest.
Compare
HeyMilo against its nearest-scoring peers.
The closest scores in Screening AI, side by side.
- Side-by-sideHeyMilo vs SmartRecruiters62D
- Side-by-sideHeyMilo vs HiPeople61D
- Side-by-sideHeyMilo vs Alva Labs61D
- Side-by-sideHeyMilo vs Braintrust AIR59F
Ranking
Others rated in Screening AI
Ranked 7 of 160 by weighted total under rubric v1.0. The ordering is arithmetic on the rubric and carries no view on which tool suits a given hiring process.
- #4SmartRecruiters62D
- #5HiPeople61D
- #6Alva Labs61D
- #7HeyMiloThis profile61D
- #8Braintrust AIR59F
- #9Sense58F
Conflicts of interest
No vendor pays for placement, scoring, or removal. Casework - the consulting firm that operates this directory - provides paid services to some vendors. Any active or recent (within 24 months) commercial relationship is disclosed on the affected vendor profile and the review is reassigned to an independent reviewer. See the full policy on About.
Casework has no commercial relationship with this vendor.