Responsible AI built for
enterprise trust and
accountability.
Olive Branch builds AI-assisted products (AIRA, Sentio, and Canopy) used by enterprises in hiring, customer experience, and workforce management. Because these products can inform decisions that affect real people, we design, test, and operate them under strict ethical and verifiable practices.
Our CLEAR AI Framework
Olive Branch's approach to responsible AI is organized around our CLEAR AI Framework: Controls, Legibility, Equity, Accountability, and Review. CLEAR is intended to provide a practical framework for developing, deploying, and operating AI Features responsibly as regulatory and enterprise requirements evolve.
Controls
OVERSIGHT & SAFETY GUARDS
- Designed to support human oversight and controlled, non-autonomous use.
- Features assist users rather than independently making consequential employment or CX decisions.
- Customers configure and maintain necessary human approval loops.
- Administrative, technical, and organizational data restriction controls.
Legibility
INTERPRETABILITY & SIGNALS
- Outputs are designed to be sufficiently understandable for human evaluation.
- Where practicable, scores or rankings explain contributing factors and signals.
- Clear documentation detailing intended purpose, inputs, and known limitations.
Equity
JOB RELEVANCE & BIAS REVIEWS
- Relies exclusively on job-relevant or business-relevant criteria.
- Routine model review for potential bias and disparate impact during development.
- Models are not designed to use protected characteristics as decision inputs unless legally required.
Accountability
CROSS-FUNCTIONAL GOVERNANCE
- Cross-functional review across Product, Engineering, Compliance, and Security.
- Formal escalation workflows for fairness, security, or regulatory concerns.
- Rigorous vetting of third-party foundation model and ML providers.
Review
AUDITING & RECALIBRATION
- Pre-deployment testing for material changes to model pipelines.
- Periodic reassessment of governance rules against emerging state and global AI laws.
- Engagement of independent third parties for model testing and validation.
Human Oversight,
Not Automated Decisions
Olive Branch's AI Features, including AIRA's candidate scoring, Sentio's conversation analytics, and Canopy's workforce insights, are strictly decision-support tools. They surface information, recommendations, summaries, and predictive scores for a human user to review and evaluate.
Core Commitment: Olive Branch does not execute automated hiring rejections, disciplinary triggers, or compensation determinations without mandatory customer human adjudication.
No Final Consequential Determinations by Olive Branch
Olive Branch does not make final hiring, disciplinary, compensation, payroll, or termination decisions on behalf of any customer. All ultimate actions are executed solely by authorized enterprise stakeholders.
Customer Control and Review Obligations
Customers configure and control how AI-generated outputs are embedded into their organizational workflows. Customers are responsible for ensuring that a qualified human reviews AI outputs before any consequential decision regarding an individual is finalized.
Contestability & Regulatory Notice Compliance
Where applicable law requires a specific level of human review, advance notice, or an opportunity to contest an automated output (for example, in high-stakes hiring or employment contexts), the customer is responsible for configuring the Services to meet that requirement. Olive Branch provides flexible product configurations and comprehensive technical documentation to facilitate this compliance.
Fairness and Bias Mitigation
We have designed AIRA and our other AI Features with the explicit operational goal of evaluating candidates, customer conversations, and workforce data purely on job-relevant and business-relevant criteria.
No Protected Characteristics as Decision Inputs
We do not intentionally design our models to utilize protected characteristics (such as race, sex, age, disability, sexual orientation, or similar categories) as decision-making inputs, except where a customer's lawful, documented affirmative program requires it and applicable law expressly permits.
Foreseeable Bias Identification & Testing
We proactively seek to identify and mitigate foreseeable sources of algorithmic bias through deliberate model architecture, rigorous multi-demographic testing datasets, and continuous post-deployment controls.
Support for Bias Audits & Local Laws (e.g., NYC LL144)
Where appropriate, we work with enterprise customers to support compliance with bias-audit and algorithmic accountability laws applicable to Automated Employment Decision Tools (AEDTs). We furnish data summaries and system documentation reasonably necessary for customer audits.
Audit Responsibility & Realistic Representations
Customers remain responsible for commissioning and publishing any statutory independent bias audit required by their jurisdiction. Olive Branch does not make warranties that any AI Feature is completely bias-free or that evaluations establish automatic compliance with all local laws.
Transparency and
Explainability
Enterprise decision-makers and candidates deserve clarity, not black-box predictions. We believe opacity is the greatest barrier to sustainable AI adoption in human capital and customer operations.
Direct match with certified enterprise architectural roles
Evaluated across past 6 years of cross-functional team execution
Age, Gender, Geographic proxies strictly zero-weighted
Data Used to Build and Operate Our Models
Data integrity and customer confidentiality are core to enterprise survival. We maintain a non-negotiable boundary between customer private datasets and foundation intelligence.
No General AI Training on Customer Data
Unless a customer expressly agrees otherwise in writing, Olive Branch does not use a customer's identifiable Customer Data to train general-purpose AI models or make that data available to improve models used by other, unrelated customers.
Aggregated & De-identified Analytics
We may use aggregated, statistical, or de-identified information to develop, test, monitor, and improve our Services, consistent with our Privacy Policy and applicable contractual obligations.
Vetted Third-Party Model Providers
Where Olive Branch uses third-party AI, machine-learning, or model providers, we evaluate provider data-use terms, privacy protections, security, and retention commitments to ensure strict alignment with our enterprise standards.
Testing, Monitoring and Change Management
AI systems require continuous validation. We maintain an engineering discipline that tests models before deployment, monitors them continuously in production, and manages changes with full audit trails.
Pre-Deployment Validation
Every model update undergoes regression testing, adversarial red-teaming, and bias auditing against benchmark datasets before reaching production. Staging environments mirror production data shapes without using real PII.
Review validation criteriaContinuous Production Monitoring
Automated monitoring tracks model accuracy, latency, error rates, and output distribution drift in real time. Discrepancies trigger alerts to our engineering and ethics teams for immediate triage.
See monitoring metricsVersioned Change Management
All models, prompts, weights, and configuration parameters are version-controlled with immutable logs. Customers can inspect version histories, rollback changes if needed, and review changelogs for any model update.
Inspect versioning processAI Governance, Accountability and Independent Assessment
Responsible AI is governed through an integrated, cross-functional framework comprising product leaders, machine learning engineers, compliance attorneys, and enterprise security personnel.
Meets quarterly and on an ad-hoc basis for material model adjustments or high-impact risk reviews.
Limitations
Olive Branch does not guarantee a particular hiring, employment, payroll, or customer-experience outcome, and does not warrant that AI outputs are free of bias, error, hallucination, security risk, or other defect, as further described in our Terms & Conditions.
Responsible use of our AI Features requires informed human judgment; we design our products, documentation, and customer onboarding to support that judgment rather than substitute for it.
Our Ongoing Commitment
Responsible AI is an ongoing program, not a one-time claim. Olive Branch is continuing to formalize and mature its AI governance practices, including internal policy documentation, risk-based model review procedures, evaluation and monitoring practices, and consideration of recognized frameworks such as the NIST AI Risk Management Framework (NIST AI RMF).
We will update this page as our program, products, customer use cases, and applicable legal requirements evolve.
NIST AI RMF Alignment
Govern • Map • Measure • Manage
Enterprise Maturity Standard
Scope and Disclaimer
This Responsible AI Disclosure describes Olive Branch's current approach to responsible AI and is provided for transparency and informational purposes. It does not constitute a warranty, certification, independent audit opinion, legal advice, or representation that any particular AI Feature is free from error, bias, security risk, or regulatory risk.
Specific contractual commitments, data-processing obligations, service descriptions, and customer responsibilities are governed by the applicable Customer Agreement, Data Processing Agreement (DPA), Terms & Conditions, and other applicable documentation executed directly between Olive Branch Technologies LLC and each customer.
Deploy responsible AI for your enterprise workforce today.
Speak with our compliance and engineering leaders about model safety, integration workflows, or requesting audit support documents.