Core Values
Integrity. Excellence. Innovation.
These are not aspirational. They are how we evaluate decisions.
Integrity
Doing the right thing, especially when it costs something. Honest pricing. Audit trails on every cluster decision. Disclosing AI involvement by default. Reporting our own mistakes to clients within hours.
Excellence
Quality work that meets the highest standard — not the minimum that gets paid. Eval-gated deployments. Defined accuracy thresholds. Conservative “if unsure, don’t” risk posture.
Innovation
Building the future responsibly. Reaching for new capabilities when they meaningfully improve service for clients — not for novelty. Staying ahead of regulation.
Non-negotiable
Our three red lines
Three commitments we will not cross, even if it costs us business.
- No deepfake voices, no impersonation of real humans. AI is always disclosed as AI. The cluster never claims to be a named person.
- No employee surveillance. Our clusters do not become monitoring tools for the people who work at our clients.
- No autonomous denials. AI may flag, recommend, or escalate, but it never autonomously denies service or cancels a customer relationship.
In Practice
Seven commitments we keep
The principles below translate our values into specific commitments. They apply to every Santiva cluster, every deployment, and every customer.
1
Humans first.
Our mission is to empower humans to achieve their full potential. Every cluster we deploy is built on one filter: does this make humans more capable without making them less essential?
2
Transparency by default.
AI involvement is disclosed in every customer-facing communication. Every cluster decision is recorded in an audit trail the client owner can inspect. Clients receive plain-language explanations of how the cluster works.
3
A path to a human, always.
Every customer the cluster interacts with has a clear way to reach a real person. We honor accuracy thresholds — 95% for customer-facing work, 99% for compliance and financial work — and escalate when we’re uncertain.
4
Privacy is the client’s, not ours.
Client data belongs to the client. Strong defaults, clear retention, full structured export at offboarding, and plain-language disclosure of what is and is not shared. Anonymized aggregate retention is disclosed at signing with an opt-out path.
5
Honest business.
Transparent pricing anchored to real operational cost. No hidden fees. No kickbacks or facilitation payments. Annual contracts with 30-day non-renewal notice. Default to swappable vendors — we will not lock clients into a stack we cannot move them off of.
6
Accountability for mistakes.
When the cluster makes an error that affects a client or an end customer, we notify the client owner within hours and the end customer within 24 hours of detection. We document root cause and remediation in writing. We do not hide mistakes.
7
Conservative by design.
We exclude industries where AI deployment carries unacceptable risk: cannabis (federally illegal), predatory lending, gambling, and adult entertainment. We monitor disclosure obligations (California SB 1001/SB 243, EU AI Act Article 50, FTC Section 5, and successors) and stay ahead of regulation.