Independent bias audits.
Evidence you can verify.
HAIEC combines qualified NYC LL144 bias-audit services, compliance software for auditors, and enterprise monitoring. From statistical analysis to signed findings, deterministic evidence packages and public verification, bring the audit and its proof together.
Audit engagements are pre-booked. Submit an availability request with your preferred timeframe. Our team confirms capacity and scope before any booking. Software can also support your ongoing audit program.
One audit. Traceable supporting evidence.
Calculation, methodology, independent professional review and publication record.
For HR leaders, independent auditors and enterprise AI teams.
Choose the way you want HAIEC to work for you. The calculation and evidence principles remain consistent, while the work, ownership and engagement differ.
Select your solution
A scoped independent audit, professionally delivered.
For employers and HR teams that want a qualified bias-audit engagement, documented calculations and publication-ready outputs without assembling their own audit operations.
- AEDT boundary, data readiness and analysis method
- Independent professional review and signed findings when issued
- Versioned calculations, evidence package and public summary support
Auditor-reviewed findings with traceable evidence.
HAIEC scopes the engagement and checks professional availability. Employer remains responsible for notices and approved public disclosures, with HAIEC assistance under agreed scope.
The audit record, the evidence package and the public reference.
Three related objects serve different purposes. Together they help HR, legal, procurement and auditors inspect what was calculated, which source and method supported it, and which record can be verified.
Explore a HAIEC evidence dossier
A labeled example showing document structure and verification meaning
Selection rate comparison
Demonstration values for a single group comparison. An issued statutory report also requires the applicable sex, race/ethnicity and intersectional analyses, data-source details and disclosed limitations.
| Group | Selected | Considered | Rate | Impact ratio |
|---|---|---|---|---|
| Reference group | 50 | 100 | 50% | 1.00 |
| Comparison group | 34 | 100 | 34% | 0.68 |
Deterministic Audit Evidence Package
Qualified inputs, audit period, methodology and rule version, normalized evidence, exclusions and calculated output are bound to an inspectable run specification and SHA-256 digests.
Verify by Artifact ID
An issued HAIEC artifact can carry a unique reference that a reviewer enters into the public verification interface. The resulting record can report integrity and status such as valid, expired, revoked or tampered.
Open HAIEC Verification Portal ↗No artifact identifier or verification status is fabricated in this demonstration. The Artifact ID is HAIEC-issued, not a New York City certification number.
The numbers should be reproducible, not mysterious.
Change the example selection counts to inspect the impact-ratio calculation. This educational calculator demonstrates why the denominator, comparison group and underlying data must be retained with the finding.
The four-fifths ratio can signal a disparity worthy of review. It does not by itself certify LL144 compliance or determine employment discrimination.
Explore HAIEC's methodology research ↗Illustrative ratio below 0.80. Review the dataset, groups and full methodology.
One annual audit is a point in time. Build an evidence trail between audits.
Use HAIEC software for statistical monitoring, change analysis and audit-preparation evidence across the authorized ATS and hiring decision flows your organization connects.
Review selection-rate changes over time
This graph illustrates the kind of trends to inspect. It is not a live enterprise monitoring feed or a claim that a finding was detected.
Decision data from your HR stack
CSV, REST APIs and supported webhooks for Greenhouse, Lever and Workday, with implementation-specific source and scope checks.
Review signals and evidence coverage
Track supported changes in statistical indicators, missing categories and data readiness across review periods.
Prepare renewals and professional review
Preserve versioned evidence, route issues for accountable review, and support the next scoped independent annual audit.
A dedicated software demonstration for your audit workflow.
See evidence intake, repeatable calculations, audit packet preparation and the boundary between monitoring signals and independent professional findings.
Three duties. One connected evidence program.
The regulatory requirements belong to the employer or employment agency using a covered automated employment decision tool. HAIEC provides software and professional services for the applicable workflow.
Independent bias audit
A covered AEDT must be subjected to an independent bias audit within one year before its use, using the applicable selection/scoring and impact ratio calculations.
Advance candidate notice
Employers and agencies must provide the prescribed notice at least 10 business days before the covered AEDT is used, with required information and request procedures.
Public Summary of Results
Publish the required statistical summary and source/audit details through the employer's appropriate public employment website, subject to the rule's requirements.
Source: NYC Department of Consumer and Worker Protection. The required public Summary of Results and HAIEC's Public Evidence Reference are separate records.
From the first scope question to evidence your reviewer can use.
Service capacity is planned in advance. We confirm availability, data readiness and engagement scope before assigning the review and delivery window.
Availability and scope
Check requested window, AEDT versions, job decisions and responsible employer.
Data qualification
Agree the data source, demographic categories, exclusions and usable periods.
Deterministic analysis
Calculate required rates and impact ratios using a documented versioned method.
Professional review
Review findings, unknowns and independence conditions; issue signed artifacts where applicable.
Publish and sustain
Support the Summary of Results, evidence verification and audit renewal plan.
Reproducibility is an assurance advantage.
HAIEC's published technical paper explains deterministic bias detection in job descriptions, version-controlled methods and verifiable records. It complements the formal statistical audit workflow rather than replacing it.
Read the Published Bias Paper ↗US-based assurance expertise · Former Fortune 100 experience · Published technical methodology (2025)
Deterministic Bias Detection for NYC Local Law 144: Why Reproducibility Matters More Than Accuracy
Versioned rules · documented behavior · cryptographic evidence
Tell us what you need. We will confirm the path.
Audit services are pre-booked, with capacity confirmed after request review. Software and continuous monitoring demos are arranged around your team's actual evaluation needs.
For audit service requests
Share your AEDT, audit type, data readiness and preferred window. Submitting a request is not a confirmed booking.
For auditors and enterprises
Tell us whether you want multi-client audit software or continuous monitoring for a connected hiring environment. We will tailor the demonstration.
Answers for HR, counsel, audit firms and enterprise teams.
Practical distinctions between the annual independent audit, software operations, ongoing monitoring and public record verification.
How do audit availability requests work?
HAIEC scopes and pre-books professional bias-audit engagements. Your form requests a preferred timeframe, but it does not reserve a date. Availability and scope are confirmed after review.
Is a software calculation the independent LL144 audit?
No. The software supplies calculations, evidence and workflow tools. The formal independent bias audit is a separately reviewed professional engagement with its own findings and scope.
Can independent audit firms use HAIEC software?
Yes. Audit practitioners can request a tailored demonstration of statistical calculations, evidence packages, multi-engagement workflow support and artifact verification. Availability of particular firm-scale features is confirmed in the demo.
Does continuous monitoring replace the annual audit?
No. Monitoring helps track signals and evidence between reviews. It does not replace the LL144 requirement for a qualifying independent bias audit within one year before use.
What exactly does the public Artifact ID verify?
A HAIEC-issued evidence identifier can support checking the status and integrity of the issued artifact. It is not a NYC government license, proof that every source record was accurate, or a substitute for the employer's public Summary of Results.
Which hiring platforms can connect?
The product includes documented Greenhouse, Lever and Workday webhook adapters as well as generic webhook, API and CSV paths. Specific setup, data mapping and source permissions are scoped to your environment.
What if our candidate data is incomplete?
Data readiness is assessed with the engagement. Missing demographic groups, exclusions, source limitations and insufficient sample data should be recorded rather than hidden behind a favorable ratio.
What does deterministic mean in a bias audit?
Accepted fixed data, calculation configuration, analysis date and methodology version support reproducible scoped calculations and evidence digests. It does not mean the underlying hiring AI is deterministic or automatically fair.