Documentation

Learn the concepts and tools behind Pod-OS

How Do I Deploy an Evolutionary Neural Memory?

Deploying an ENM is done through the Pod-OS Management Dashboard. An ENM is an Actor, so deploying one follows the same workflow as deploying any other Actor in your network.

Prerequisites

Before deploying an ENM, make sure you have:

  1. An active Pod-OS deployment with at least one Gateway running
  2. A user account with the developer, operator, tenant_administrator, or tenant_owner role
  3. Access to the Pod-OS dashboard

Step-by-Step Deployment

1. Navigate to the Dashboard

Log in to the Pod-OS Management Dashboard and go to the Dashboard page from the left sidebar.

2. Deploy an Actor Network

If you do not yet have an actor network, click Deploy Actor Network in the Quick Actions section. This will walk you through:

  • Naming your Gateway
  • Selecting a subscription level (which determines available resources)
  • Choosing which Actor types to include in your deployment

3. Add an ENM Actor

When configuring your actor network, select Evolutionary Neural Memory as one of the Actor types. You will need to provide:

Field Description
Actor Name A unique name for this ENM instance (e.g., product-catalog, user-profiles)

4. Deploy

Click Deploy to begin the deployment process. Pod-OS will:

  1. Create the Kubernetes resources for your ENM
  2. Register the ENM as an Actor in the actor network
  3. Configure routing so your Gateway can manage message traffic
  4. Run health checks to verify the ENM is operational

You can monitor the deployment progress in real-time through the dashboard's streaming status display.

5. Verify

Once deployment completes, your ENM will appear in the Actors page in the sidebar. You can:

  • View its status and health
  • See its network connections
  • Send test messages via the Direct Messaging tool

Managing Your ENM

After deployment, you can manage your ENM through the dashboard:

  • Actors page — view all Actors including your ENMs, inspect their status, and edit configuration
  • Neural Memory Explorer — query and explore data stored in your ENMs
  • Direct Messaging — send raw messages (on-the-wire socket messages) to your ENM for testing and debugging

Scaling Considerations

Each ENM is a self-contained Actor. To scale:

  • Vertically — increase the resources allocated to the ENM's pod through Gateway configuration
  • Horizontally — deploy additional ENM instances for different data domains

Because ENMs are share-nothing, there is no coordination overhead between instances. Each one operates independently.

Next: How do I add data to an ENM? — learn how to ingest data using the Pod-OS SDKs.