Terminology

## Platform-Specific Terminology

**hosted·ai User Panel**: The customer-facing interface for end users to manage their resources and teams. Service providers also use this to customize offerings and manage resources and teams.

**hosted·ai Admin Panel**: A single pane of glass for service providers to manage infrastructure recipes and GPUaaS Pools across all regions.

**hosted·ai Cluster Panel**: An interface for service providers to configure and manage infrastructure in a single region. It provides direct control over physical nodes, networks, storage, and virtualization resources.

**hosted·ai User API**: Programmatic interface for end users to manage their resources and teams via API. Service providers can also use this API to customize offerings and manage resources and teams programmatically.

**hosted·ai Admin API**: API for service providers to manage infrastructure recipes and GPUaaS Pools across all regions. The User Panel also uses this API to perform internal operations.

**hosted·ai Cluster API**: API for service providers to configure and manage infrastructure in a single region. The Admin Panel uses this API to perform internal operations. It provides low-level control over physical resources, virtualization, and cluster-specific configurations.

**hosted·ai KVM Node**: Infrastructure component that provides CPU, Memory, Storage (and optionally GPU) to be consumed by virtual machine instances. This can include virtualized GPUaaS Nodes. (Coming soon)

**hosted·ai GPUaaS Node**: A virtual machine or bare metal server that can have multiple roles (Worker, Controller, Storage, Gateway) in the GPU-as-a-Service architecture.

## A

**Admin Panel**: The administrative interface used by service providers to manage the underlying infrastructure, configure resources, and monitor the platform.

**API Keys**: Secure authentication tokens that allow programmatic access to the HostedAI platform's features and resources.

**Audit Logs**: Records of actions performed within the platform, tracking who did what and when for security and compliance purposes.

## B

**Billing**: The process of calculating charges based on resource usage and subscription plans.

## C

**Cluster**: A group of interconnected servers that work together as a unified computing resource.

**Cluster Resource Group**: A logical grouping of computing resources within a cluster that can be allocated to users.

**Customer Portal**: See "User Panel."

## G

**GPU-as-a-Service (GPUaaS)**: A cloud service model that provides on-demand access to GPU computing resources without requiring customers to own physical hardware.

## H

**HAIShare**: The hosted.ai scheduler which ensure GPU workloads run at the right place, at the right time based on the workload and pool characteristics.

## I

**Infrastructure**: The underlying hardware, software, networks, and facilities that support the HostedAI platform.

**Instance**: A virtual machine or computing environment provisioned for a customer's use.

**Instance Type**: A predefined configuration of computing resources (CPU, memory, storage) available for provisioning.

## N

**Network**: The communication infrastructure that connects computing resources and enables data transfer between them.

## P

**Permissions**: Rules that determine what actions users can perform within the platform.

**Policies**: Sets of rules that govern resource allocation, usage limits, and security requirements.

**Provisioning**: The process of setting up and configuring computing resources for customer use.

## R

**RBAC (Role-Based Access Control)**: A method of regulating access to resources based on the roles assigned to users.

**Recipe**: A predefined template for creating and configuring resources with specific settings.

**Region**: A geographic area where HostedAI resources are located.

**Resource**: Any computing asset (such as compute instances, storage, or network components) that can be provisioned and used by customers.

**Resource Allocation**: The process of assigning computing resources to specific users or workloads.

## S

**Service Provider**: The organization that operates the HostedAI platform and provides cloud services to customers.

**SSH Keys**: Cryptographic keys used for secure authentication when accessing instances via SSH (Secure Shell).

**Storage Block**: A unit of persistent storage that can be attached to computing instances.

**Subscription**: A payment plan that determines the level of access and resources available to a customer.

## T

**Team**: A group of users who collaborate and share access to resources within the HostedAI platform.

## U

**User Panel**: The customer-facing interface where end users manage their resources, team members, and billing information.

## W

**Workspace**: A logical container for organizing related resources and their costs.