LiyaSupport
Security & Trust

Built for teams that
can't afford surprises

LiyaSupport is designed so teams can safely adopt AI without sacrificing visibility, governance, or control.

SOC 2 Ready
GDPR Aligned
SSO / SAML
Immutable Audit Logs
Data Residency
Role-Based Access

Compliance & Governance

Enterprise controls, ready on day one

The compliance features regulated industries require — without the 6-month implementation project.

SOC 2 Ready

Security controls aligned to SOC 2 Type II requirements across security, availability, and confidentiality trust service criteria.

GDPR Aligned

Data processing agreements, right-to-erasure workflows, and data residency controls built in for EU customers.

Immutable Audit Logs

Every ticket action, AI draft, agent edit, and send is recorded with full attribution — who did what, when, and how.

Role-Based Access Control

Granular permissions for Admin, Agent, and Read-only roles. Restrict access to sensitive queues or customer data by role.

SSO / SAML

Native support for SAML 2.0 SSO via Okta, Azure AD, Google Workspace, and any compliant identity provider.

Data Residency

Choose where your support data is stored. EU and US regions available, with region-locked tenant isolation.

AI Governance

AI that works
with your team, not around it

Every AI response is grounded, scored, and reviewed by a human before it reaches a customer. There are no autonomous sends.

Our AI guarantee

Liya never sends a message without agent approval. AI drafts. Humans decide.

AI Quality Monitoring

Every AI-drafted response is scored for accuracy, tone, and grounding before it reaches an agent. Low-confidence responses are automatically flagged.

Knowledge Grounding

AI responses are grounded in your help docs, macros, and verified answers — not open-ended generation. Hallucinations are structurally prevented.

Flagged Response Reviews

Responses below your confidence threshold enter a review queue before they can be sent. Agents approve, edit, or reject — the AI never sends autonomously.

Human-in-the-Loop

Every response requires an agent to review and send. Liya prepares and assists — your team always has final say before a customer receives a message.

Data Protection

Your data stays yours

Customer conversations, knowledge bases, and support history are encrypted, isolated, and never used to train shared models.

Encryption in Transit

All data transmitted between your team, Liya, and your customers is encrypted using TLS 1.2+ with HSTS enforcement.

Encryption at Rest

Customer data and conversation history are encrypted at rest using AES-256. Encryption keys are managed per-tenant.

Secure Knowledge Storage

Your help docs and macros are stored in isolated, encrypted indexes. They are never used to train shared models.

Tenant Isolation

Every customer's data is fully isolated at the infrastructure level. There is no shared storage or cross-tenant data access.

Safe AI adoption starts with the right platform

LiyaSupport is designed so teams can safely adopt AI without sacrificing visibility, governance, or control.