Welcome to the RI Platform!

The Robust Intelligence (RI) Platform enables comprehensive testing of data and model performance at the pre-development and post-development stages of the model life cycle. Users already familiar with the RI Platform who want to know what’s new in the most recent release can find details in the Release Notes.

• Release Notes

The RI Platform documentation covers several broad topic areas. If you’re new to the RI Platform or want to familiarize yourself with core concepts of machine learning, Getting Started provides an overview, as well as a comprehensive walkthrough of a local trial of the RI Platform.

Getting Started

• RIME overview
  • The RI Platform
  • Why use the RI Platform?
  • Key Features
  • Key Machine Learning Tasks Covered
  • RI Platform Deployment Patterns
  • Summary Tests
  • RI Platform consolidated test database
  • Tabular
  • NLP
  • CV
  • RIME local trial

The Administering the RI Platform section discusses tasks related to the maintenance and administration of an RI Platform instance, including user account management, role assignments, authentication systems, and API access.

Administering the RI Platform

• Administering the RI Platform
  • Managing users
  • User account roles
  • Role-based Access Control
  • Native user authentication
  • SSO configuration
  • API access tokens
  • Workspaces
  • Managing Workspaces
  • Managing agents
  • Managed Images
  • Configuring data sources
  • Audit Logs
  • Configuring SMTP for alerts
  • FAQ

RI Platform instances are available as cloud installations, either fully managed by Robust Intelligence or directly managed by you. The RI Platform can also be installed as a stand-alone, self-hosted application. The installation process is discussed in Installing the RI Platform.

Installing the RI Platform

• Installation Guide
  • Managed Cloud
  • Self-Hosted
  • Cloud

The Data science with the RI Platform section discusses actively using the RI Platform to evaluate data sets and models. Configuring alerts, setting up integrations, and running tests on uploaded models are all discussed here.

Data science with the RI Platform

• Get Started
  • Setup
  • Tabular Notebooks
  • NLP Notebooks
  • CV Notebooks
  • Custom Notebooks
• How-To Guides
  • API Authentication
  • NLP Adversarial Robustness
  • Comparing Two Test Runs
  • Compliance
  • Uploading Data to RIME
  • Querying RIME Results
  • Updating a Firewall
  • Using the Data Collector
  • Scheduling Continuous Testing Runs
  • Using the REST API
  • Model cards
  • Custom embeddings
  • Projects
  • Platform Integrations
  • Model Integrations
  • Data Integrations
  • Alerts and Notifications
  • Test Run Progress
  • Custom Metrics
  • Custom Tests
  • Run RIME on Managed Docker Images
  • Debugging the Python SDK
  • RIME Data Format Check
• Reference
  • Tabular Configuration
  • NLP Configuration
  • CV Configuration

The Reference section contains in-depth details about the Python SDK, including detailed JSON templates to configure stress testing for the RI Platform.

Reference

• RIME in-depth reference
  • RIME Python Library