MaaS Platform Documentation

Welcome to the Models-as-a-Service (MaaS) Platform documentation.

The MaaS Platform enhances the model serving capabilities of Open Data Hub by adding a management layer for self-service access control, rate limiting, and subscription-based entitlements.

Use this platform to streamline the deployment of your models, monitor usage, and effectively manage costs.

📚 Documentation Overview

🚀 Getting Started

• QuickStart Guide - Complete platform deployment instructions
• Architecture - Overview of the MaaS Platform architecture

👤 User Guide

• API Key Management - Creating and managing API keys
• Model Discovery - Listing available models
• Inference - Making inference requests

⚙️ Administration

• Access and Quota Overview - Configuring policies and subscriptions
• Model Setup - Setting up models for MaaS
• API Key Administration - Bulk revocation and cleanup
• Observability - Monitoring, metrics, and dashboards
• Limitador Persistence - Redis backend for rate-limit counters
• TLS Configuration - Configuring TLS

📋 Release Notes

• Release notes - Version highlights and known limitations by release

📖 Installation Guide

• Prerequisites - Requirements, database setup, observability and GenAI Studio prerequisites
• Platform Setup - Install ODH/RHOAI with MaaS
• MaaS Setup - Gateway, DataScienceCluster, OdhDashboardConfig feature flags
• Validation - Verify your deployment

🔄 Migration

• Tier to Subscription Migration - Migrate from tier-based to subscription-based access control