Security

Define who can act. Control what can leave.

Connect your identity provider, scope access to models and tools, and apply data policies around the conversation. Kaldryn puts these controls in the same console as the AI they protect.

Explore this part of the platform

Start with identity

Configure SAML metadata, allowed email domains and attribute mapping. Just-in-time provisioning and group-to-role mapping connect existing identities to portal roles.

Constrain every capability

Grant model tiers by role. Set connector tool allowlists and timeouts, use per-user credentials where configured, and choose whether data rules block, redact or audit a message.

Verify the boundary

Inspect destination-level egress probes, isolation checks and audit events. See the difference between a private service, a blocked public endpoint and an unresolved destination.

Explore this part of the platform

Kaldryn Platform · Security

Interactive software preview · illustrative data

Choose an admin page

DLP & Policies

Govern sensitive information before it becomes an incident.

Illustrative data · safe to explore

Rules

Redacted

Blocked

Policy test

Test a sample message

This sample checks email patterns only. The product supports additional configurable rules.

Custom pattern rules

Preview only · no changes to your systems

What this module includes

Block, redact or audit
Sensitivity labels and information barriers
eDiscovery and department policies
Custom pattern rules
Incoming and outgoing message policies
Rule severity and violation review
Department-to-department barriers

Overview

Data governance, in one view

Personal data · Redact

Credentials & secrets · Block

Policy events · Audit

DLP Events

Review what the policy engine caught

Sensitive prompt · Redacted

Credential pattern · Blocked

Event review · Assigned

DLP Rules

Choose the response to sensitive data

Personal identifiers · Redact

API keys · Block

Custom patterns · Audit

Violations

Investigate and resolve policy violations

Event context · Review

Policy owner · Assign

Resolution · Record

eDiscovery

Find the evidence behind an answer

Conversations · Search

Users & date range · Filter

Review results · Export

Barriers

Keep restricted knowledge in the right teams

Legal / Operations · Restricted

Confidential libraries · Scoped

Cross-team access · Policy-controlled

Policies

Turn governance into enforceable rules

Sensitivity · Classify

Access · Control

Activity · Audit

Departments

Apply policies in your organisational context

Legal · Department policy

Finance · Department policy

Operations · Department policy

Based on the Kaldryn Platform admin console. Available features depend on license tier, role, hardware and configuration.

Network isolation depends on deployment and firewall configuration. The console exposes checks and evidence; enabling a feature alone does not guarantee isolation.

Inside these modules

User ManagementGive people the access their work requires.Explore capabilities

Invitations and account lifecycle
Roles and departments
Authentication and account controls
Activity and risk indicators
API-key rotation
Account activation and deactivation
Password-reset links

SSO / SAMLConnect enterprise identity to the AI platform.Explore capabilities

SAML identity-provider configuration
SSO sign-in
Role-aware access to the console
Import IdP metadata XML
Just-in-time user provisioning
Allowed email domains
Attribute and group-to-role mappings

Model AccessControl which users can access which models.Explore capabilities

Per-model permissions
User and role access rules
Central policy administration
Quick, Mid and Deep Analysis tiers
CPU and vision access
Auto-routing permissions
Role-specific model configuration

ConnectorsConnect internal systems through a governed MCP tool layer.Explore capabilities

Connector registration and health
Per-user identity mappings
Tool-call audit and DLP controls
Inspect available MCP tools
Latency and connection tests
Per-user downstream credentials
DLP, rate-limit and timeout outcomes

DLP & PoliciesGovern sensitive information before it becomes an incident.Explore capabilities

Block, redact or audit
Sensitivity labels and information barriers
eDiscovery and department policies
Custom pattern rules
Incoming and outgoing message policies
Rule severity and violation review
Department-to-department barriers

Model TrustMake model selection a documented security decision.Explore capabilities

Model provenance and trust inventory
Trust policy administration
Security review evidence
Minimum model-trust threshold
Permissive to strict policy slider
Installation outcome history
Review recorded overrides

Discuss your deployment