AI Governance

1. Purpose of AI Systems

NSigma develops AI systems designed to assist organizations with analytics, forecasting, and decision-support. These systems are intended to augment human judgment rather than replace it.

Our AI solutions are purpose-built for specific domains including investment management, commercial real estate, utilities and infrastructure, and waste management — delivering measurable value through data-driven insights.

2. Data Usage Policy

Client-provided data is not used to train general-purpose models unless explicitly authorized by the client.

NSigma maintains strict data boundaries to ensure:

• Client data remains isolated and is processed solely for the agreed-upon purpose
• No cross-client data sharing occurs without explicit written consent
• Data retention and deletion policies are enforced in accordance with contractual agreements
• All data handling complies with applicable privacy and data protection regulations

3. Model Transparency

NSigma AI systems may produce probabilistic outputs based on statistical models and machine learning algorithms. Outputs should be interpreted as analytical insights rather than deterministic conclusions.

We are committed to transparency in how our models operate:

• Model methodologies and limitations are documented and communicated to clients
• Confidence levels and uncertainty ranges are provided where applicable
• Assumptions underlying model design are disclosed as part of engagement documentation

4. Human Oversight

AI-generated insights are intended to support human decision-making and should be evaluated alongside domain expertise and independent analysis.

NSigma designs its systems with human-in-the-loop principles:

• Critical decisions are not delegated solely to automated systems
• Users retain the ability to review, override, and contextualize AI outputs
• Escalation pathways are built into workflows for edge cases and anomalies

5. Model Monitoring

Models are periodically evaluated for performance, drift, and reliability to ensure outputs remain relevant and accurate.

Our monitoring framework includes:

• Continuous performance tracking against established benchmarks
• Drift detection to identify changes in data distributions or model behavior
• Scheduled retraining and recalibration cycles
• Incident response protocols for unexpected model degradation

6. Ethical Use

NSigma does not design AI systems for unlawful, discriminatory, or harmful uses and expects clients to deploy systems responsibly.

Our ethical commitments include:

• Designing systems that respect fairness and avoid unjust bias
• Refusing engagements that conflict with ethical standards
• Supporting clients in establishing responsible AI deployment practices
• Ongoing review of our AI practices against evolving industry standards and regulations

Contact

For questions about NSigma's AI governance practices, please contact: