System Architecture
MeridianAI Intelligence Stack
Seven layers from secure deployment to the interactive experience layer. Every autonomous action is logged, sandboxed, permissioned, and reviewable.
01
Experience Layer
Mission dashboardResearch workspaceScenario labAgent builderOutput center
02
Secure API Layer
AuthenticationAuthorizationRESTgRPCIntegrationsWebhooks
03
Agent Runtime
PlannerOrchestratorAgent registryRecursive delegationTask ledgerEvaluation
04
Model Intelligence Layer
Language modelsSpecialist ML modelsSimulation enginesOptimization systemsPolicy filters
05
Knowledge Fabric
Knowledge graphRetrievalDocumentsResearch repositoriesStructured databasesTelemetryProvenance
06
Security & Governance
IdentityRBACSandboxingEncryptionAudit ledgerPolicy enforcementHuman approvals
07
Deployment Plane
ContainersKubernetesCloudHybridOn-premIsolated environments