Agent Runtime
The execution engine that runs autonomous agents against your processes. Reads your approval policies, classifies requests, routes decisions, and writes to your audit storage — without calling any external API.
Explore Agent Runtime
Platform Architecture
The dodoAI Platform
Three layers that let enterprise AI operate without cloud dependency. Agent execution, deployment control, and integration — all running inside your perimeter.
Core Layers
Three layers. One perimeter.
Each layer has a single responsibility — and all three run entirely within your infrastructure.
Deployment Engine
Handles installation, configuration, and lifecycle management of dodoAI inside your environment. Supports on-premises bare metal, VMware, private cloud VPC, and air-gapped networks. Four-week typical onboarding.
View Deployment GuideIntegration Layer
Connects dodoAI to your existing enterprise systems — ERP, procurement platforms, compliance databases, document management, and directory services — using standard protocols. No cloud relay required.
Browse Integrations
Architecture
Everything inside the boundary
dodoAI sits between your data sources and process outputs — and that boundary never has an outbound connection.
Ready to see the platform in detail?
We walk every prospect through our architecture before any deployment conversation. No commitment required.