Legible analyzes structural and behavioral metadata to understand how distributed systems operate over time.
Legible processes:
• Service interaction metadata and workflow transitions
• Deployment and version signals
• Topology and dependency relationships
• Execution state transitions
Legible does not collect:
• Application payload data
• Request or response bodies
• Customer database content
• Long-term execution traces
Our platform evaluates behavioral structure, not business data.
Legible follows the principle of least privilege.
• Read-only access to required telemetry sources
• Scoped IAM permissions
• No inline traffic interception
• No direct modification of production systems
Legible does not autonomously execute remediation actions. If customers choose to expose controlled management interfaces (e.g., MCP endpoints), any actions remain explicitly authorized and governed within the customer environment.
Operational authority always remains with engineering teams.
• Encryption in transit (TLS 1.2+)
• Encryption at rest (AES-256)
• Tenant isolation
• Role-based access controls (RBAC)
• Audit logging of platform access and configuration changes
Access to customer environments is restricted and logged.
• SAML / OIDC-based Single Sign-On
• Role-based authorization
• MFA enforcement via identity provider
• Environment-level access segmentation
Access is granted strictly according to least-privilege principles.
Legible supports deployment models suitable for complex environments with 30+ services, including:
• Secure multi-tenant SaaS
• Dedicated single-tenant environments
• Private VPC deployment options
Deployment architecture is aligned with enterprise security requirements.
Legible security practices align with SOC 2 control principles. SOC 2 Type I certification is currently in progress. We maintain internal policies covering:
• Access management
• Secure software development lifecycle (SDLC)
• Change management
• Incident response
Legible evaluates whether systems continue to operate as the intended system over time. Traditional monitoring ensures systems are available. Legible ensures system identity and behavioral integrity are preserved. The platform provides explainable outputs and deterministic validation criteria without autonomous control over production systems.
Legible technologies are protected by filed U.S. provisional patent applications and pending patent filings. See our Patent Notice page for additional information.