Systemic Integrity Integration Protocol
SIIP is enterprise compliance middleware. It is designed to sit within the HRIS environment an organization already runs and evaluate high-risk employment decisions before they reach the system of record, whether those decisions are produced by an automated tool or initiated by a person.
Your HRIS is where employment decisions are recorded. Your managers and automated tools are where those decisions originate. Today, nothing sits between the two.
SIIP is designed to occupy that position — receiving a high-risk employment decision on its way to the system of record, evaluating it against the obligations that apply, and permitting it, holding it, or directing it to review before the record is written.
Everything else on this page is a description of how that single idea is implemented.
Capability One
Automated and AI-enabled tools now contribute to screening, ranking, scoring, evaluation, compensation, and separation across the enterprise.
SIIP is designed to evaluate an automated employment decision before it takes effect, examining it for indications of protected-characteristic influence and for conflict with the employee’s current compliance state.
NYC Local Law 144, Colorado AI Act — Effective June 30, 2026, the EU AI Act, and the EEOC’s enforcement priorities all create obligations around automated employment systems.
SIIP is being developed to address the obligations these frameworks create. It does not replace a bias audit, legal counsel, or an organization’s own compliance judgment.
Capability Two
Protected leave, an open accommodation request, a recent internal complaint or regulatory filing can change the legal character of an otherwise routine action.
That protected state is usually known to the organization, recorded in a case file, leave system, or compliance log. It is rarely known to the system processing the transaction, and almost never known at the moment the transaction is entered.
SIIP is designed to close that gap: to verify an employee’s active protected status at the point a high-risk action is initiated, and to hold that action before it commits when the status and the action conflict.
The Sequence
Origin, automated or human, does not change the path.
The decision is checked against the compliance conditions that apply to it.
Every evaluation is recorded regardless of outcome.
Six-Patent Architecture
6 USPTO Provisional Patent Applications
Filed: July 8, 2026 & August 6, 2026
Application #64/107,645 (July 8) + 5 additional applications (August 6)
Operates as a persistent API-layer interception gateway, structurally positioned between the output endpoints of third-party HRIS and ATS platforms and their downstream execution environments. The Module intercepts structured employment decision data objects — including candidate scoring vectors, compensation recommendation outputs, workforce action triggers, and automated performance evaluation results — at the API response layer. Interception occurs prior to any write operation reaching the host platform's execution queue. No automated employment action proceeds to execution without first transiting this mandatory gateway.
Applies a multi-dimensional protected-class correlation engine to each intercepted employment decision data object in-process, prior to any data transmission or downstream handling. The engine computes statistical divergence coefficients across demographic variable clusters derived from the decision payload's feature vectors, executing entirely within the intercepted data object's processing context. Computed divergence values are evaluated against a pre-compiled regulatory threshold matrix — mapping each coefficient to applicable federal disparate impact standards under Title VII, the ADA, and the ADEA — and the engine outputs a structured compliance signal object. No human judgment is introduced at this stage. The output is generated entirely by the deterministic application of the threshold matrix to the computed coefficients, producing a machine-readable signal that controls all subsequent routing.
Operates as a deterministic finite-state automaton whose state-transition table is compiled from an actively maintained regulatory mandate dataset. Upon receiving the compliance signal object from the Evaluation Node, the Engine traverses its pre-computed transition paths — mapped against applicable federal, state, and municipal mandates including Title VII (1964), the Americans with Disabilities Act (1990), the Age Discrimination in Employment Act (1967), NYC Local Law 144 (enforcement: July 5, 2023), and the Colorado AI Act (effective: June 30, 2026) — and outputs a structured compliance verdict object. The verdict object contains two components: a binary routing directive controlling all downstream data flow, and a mandate-citation index identifying the specific regulatory provisions evaluated against the intercepted payload.
Operates as a deterministic data-flow controller that enforces routing directives issued by the Compliance Decision Engine. Compliant decision objects — those whose compliance verdict object carries a cleared routing directive — are released and transmitted to the originating host platform's execution endpoint, permitting the employment action to proceed. Non-compliant decision objects are intercepted, removed from the active data flow, and written to an isolated quarantine buffer, blocking all transmission to the execution endpoint and preventing the non-compliant action from reaching any downstream system, database, or human workflow. The routing enforcement is structural and automatic — no manual intervention is required or possible at this stage.
Generates a sequential, append-only ledger record for each employment decision data object processed through the architecture. Each ledger record comprises five immutable components: (1) a cryptographic hash of the original intercepted decision payload; (2) the computed compliance signal output generated by the Evaluation Node; (3) the mandate-citation index produced by the Compliance Decision Engine; (4) the routing directive enforced by the Enforcement Node; and (5) a UTC-synchronized timestamp recorded at each stage transition. Each ledger record is cryptographically chained to the immediately preceding record — rendering any post-hoc modification or retroactive insertion detectable through hash-chain verification. The resulting ledger constitutes a complete, tamper-evident sequential record of every governance action applied to every employment decision processed by the system.
Patent 4 · SIIP Carrier-Link
Real-time bidirectional data pipeline connecting SIIP's compliance engine to insurance carrier actuarial systems via the CENP (Compliance Event Normalization Protocol). Transmits structured compliance event data for live EPLI underwriting adjustment.
Patent 6 · SIIP Risk-Share
Continuous Compliance Scoring Function (CCSF) computing S(t) = clip(w₁·PR + w₂·(1−ER) + w₃·RS − w₄·PCE + w₅·TC, 0, 1) in real time. Drives ARRP and PRSA performance revenue share allocation — generating carrier-funded revenue share from every protected employment decision.
SIIP's commercial architecture is protected across six USPTO provisional patent applications.
Integration Layer
SIIP is intended to operate inline with leading HRIS platforms rather than requiring a rip-and-replace deployment.
Integration sequencing remains subject to enterprise environment review and early-access deployment planning.
Regulatory Environment
Six active regulatory mandates now shape AI employment decision risk across the United States and Europe.
Disparate impact and disparate treatment standards for AI employment tools.
SIIP: Real-time 4/5ths rule enforcement across all protected class dimensions.
Mandatory annual bias audit for AI employment decision tools. Applies to all NYC employers using automated employment decision tools.
SIIP: Continuous audit — exceeds LL144.
Direct employer liability for consequential AI employment decisions affecting protected classes. C.R.S. §24-34-606.
SIIP: Real-time intercept = direct liability shield.
Employment AI classified as high-risk. Mandatory conformity assessment, continuous monitoring, audit logging.
SIIP: Patent 5 Vault satisfies EU Article 12.
Video interview and AI hiring tool regulation. Employer notice, consent, and data retention requirements.
SIIP: Automated compliance documentation.
Federal contractors must document AI employment decision processes for OFCCP audit compliance.
SIIP: Patent 5 audit chain = OFCCP-ready.
This overview is informational and does not constitute legal advice. Coverage and obligations depend on the organization, jurisdiction, system, and use case.
Deployment
Initial integration targets are Workday, ADP, Oracle HCM and SAP SuccessFactors, which together account for the large majority of enterprise HRIS environments in the company’s target market.
The platform is being developed on Google Cloud Platform.
Designed for Accountable Leaders
For the CHRO
SIIP is designed so compliance becomes a property of the transaction rather than a matter of individual judgment.
For the Chief Compliance Officer
Each evaluation produces a live record of what was submitted, what was found, and what followed.
For the EPLI Underwriter
SIIP is designed to make prevention and governance controls visible rather than inferred.
Early Access
Participating organizations influence integration sequencing, review workflow design, and reporting output.