SDKs

Tensor One SDKs give programmers powerful tools for interacting with the Tensor One API. They facilitate the smooth integration of custom logic and simplify infrastructure management.

Interacting with Serverless Endpoints

Once deployed, serverless functions are accessible via Serverless Endpoints, which external applications can call using standard HTTP requests.

How It Works

• Serverless Endpoints behave like HTTP APIs.
• You must provide your Endpoint ID and API Key to authorize and complete requests.

This enables programmatic interaction with your deployed functions from other apps or services.

Infrastructure Management with the SDK

The Tensor One SDK allows for full programmatic control of your cloud environment, including:

• Clusters
• Templates
• Serverless Endpoints

Managing Clusters

Clusters are isolated environments used to run applications.

Key Actions:

• Create a Cluster: Instantiate new Clusters with specific hardware and software configurations.
• Configure Resources: Define GPU type, memory allocation, storage, and network settings.
• Deploy Applications: Push code or services directly to Clusters.
• Monitor and Scale: Use the SDK to track performance and automatically adjust resources.

Managing Templates and Endpoints

Templates define reusable Cluster configurations, while Endpoints expose services to the web.

Workflow:

1. Create a Template: Specify base OS, dependencies, disk size, environment variables, and port settings.
2. Launch Clusters from Templates: Easily spin up consistent environments across deployments.
3. Create Endpoints: Enable secure external access to the applications running inside Clusters.

Using Tensor One SDKs, you can fully automate deployment workflows, configure infrastructure at scale, and expose intelligent services to the world via secure, serverless APIs.