Skip to Content

isA Platform

Deploy and scale your AI agents in the cloud.

Building a governed enterprise app? Read the AEP control plane guide to understand how projects, capabilities, approvals, evidence, and releases fit together.

Deploying isA for a specific customer or brand? Read isA Editions to understand the brand-as-config model that replaced the old white-label fork approach.

Overview

The isA Platform provides infrastructure for deploying, managing, and scaling AI agents built with the Agent SDK.

Key Features

Code Deployment

Deploy your custom agents to production with a single command:

# Upload and deploy
tar -czf my-agent.tar.gz -C my-agent .
curl -X POST /api/v1/deployments/{id}/upload -F "file=@my-agent.tar.gz"
curl -X POST /api/v1/deployments/{id}/deploy

Auto-Scaling

  • Automatic scaling based on load
  • Configurable min/max replicas
  • Zero-downtime deployments

Isolation & Security

  • VM-level isolation for each execution
  • Network isolation (no direct external access)
  • Sandboxed tool execution

Monitoring

  • Real-time logs and metrics
  • Health checking
  • Alerting and notifications

Architecture

User Request → API Gateway → Pool Manager → Cloud OS VM
                                           └─→ API Proxy → External Services

Deployment Workflow

  1. Create deployment configuration
  2. Upload code (tar.gz or Git)
  3. Automatic Docker image build
  4. Deploy to Cloud OS VMs
  5. Route traffic to your agent

Roadmap

The following features are planned for the isA Platform:

  • CLI tool (isa deploy) — Deploy agents from the command line
  • CI/CD integration — GitHub Actions and GitLab CI workflows
  • Custom domains — Map your own domain to deployed agents
  • Environment variables — Inject config into agent runtime
  • Secrets management — Secure credential injection via Vault

See the Cloud deployment guide for current deployment options.

Home