On-Premises
Full deployment on your bare metal or VMware infrastructure. dodoAI installs as a service on your existing Linux servers. No cloud account required at any stage.
Platform · Layer 2
Deploy Inside Your Perimeter
Three deployment modes. One architecture principle: your data never leaves. Typical onboarding runs 4 weeks from infrastructure access to first agent live in production.
Deployment Modes
Choose your environment
All modes share the same architecture guarantee: no data egress, no external model calls.
Private Cloud VPC
Isolated tenant deployment inside your private cloud VPC. Data remains within your cloud account boundary. No shared infrastructure with other dodoAI customers.
Air-Gapped
Zero internet access required at runtime. Model weights are transferred once during onboarding. After that, dodoAI operates indefinitely without any external connectivity.
Onboarding
Four weeks to first agent live
Our onboarding process is designed around your IT team's constraints — not ours.
1
Week 1
Environment access and dependency audit
Your IT team grants limited access to the target deployment environment. We audit existing infrastructure — OS version, available resources, network topology, and firewall rules — and confirm compatibility.
2
Week 2
Agent runtime install and configuration
We deploy the dodoAI runtime inside your environment. This includes the execution engine, model weights, and integration adapter framework. No external connections made. All artifacts transferred via your approved channels.
3
Week 3
Rule mapping from your existing policies
Our team works with your operations and compliance leads to map your approval policies, vendor criteria, and compliance thresholds into agent-readable rule sets. This is the most business-critical step — we do it carefully.
4
Week 4
First agent live, monitoring enabled
First agent runs in production with your team observing. Audit log review confirms expected behavior. Monitoring dashboards configured in your own environment. Handoff documentation delivered.
Technical Requirements
Infrastructure prerequisites
Minimum requirements for a standard single-agent deployment. Multi-agent configurations scale linearly.
| Component | Minimum Specification | Notes |
|---|---|---|
| CPU | 8 cores (x86_64) | 16+ cores recommended for concurrent agent runs |
| RAM | 32 GB | 64 GB recommended for multi-process workloads |
| Storage | 200 GB SSD | Audit log storage is additional and sized per retention policy |
| OS | RHEL 8/9, Ubuntu 20.04/22.04, Debian 11/12 | Standard enterprise Linux distributions supported |
| Network | Internal LAN only | No outbound internet required at runtime. Onboarding transfer via approved channels |
| Ports (internal) | 8080 (agent API), 8443 (admin), 5432 (audit DB) | All traffic stays within your network segment |
| Containerization | Docker 24+ or Podman 4+ | Kubernetes supported but not required |
Start your deployment planning
Send us your environment details and we will confirm compatibility before any commercial conversation.