Skip to main content

Codex Remediations

Codex’s Remediations works alongside Logs to create a comprehensive AI safety system. While Logs detect potential failures, Remediations both patch critical issues and provide expert-curated answers.

Within the Remediations section of the Codex Web App or directly from failure cases in Logs. Every intervention’s impact is automatically tracked, helping you measure improvements to your AI application’s safety and reliability.

Codex Remediations Interface

How do SMEs use Remediations?​

Adding and Managing Remediations​

There are two ways to create remediations:

1. Quick Remediations from Logs​

When reviewing the Logs view, you can create remediations directly from unaddressed failure cases:

  • Click “Create Remediation” on any unaddressed log entry
  • View the full context and metadata of the original failure while creating the remediation
  • Once created, your remediation will automatically protect against similar failures
  • Codex intelligently matches your remediation to relevant future queries and responses

Quick Remediate Interface

2. Standard Remediation Management​

From the Remediations interface, you can:

  • Create new remediations (starts in Draft status)
  • Activate remediations to start protecting against failures
  • Pause remediations to temporarily disable them
  • Edit or delete remediations (only if they haven’t been used to prevent failures)
  • View all Active, Paused, and Draft remediations in one place
  • See which types of queries and responses your remediation will automatically protect

What information is tracked?​

As you work with Remediations, Codex maintains detailed records to help you monitor and improve your AI application’s safety:

1. Remediation Metadata​

Each remediation entry includes key information about its content and status:

  • Status (Active, Paused, or Draft)
    • Active: Currently protecting against failures
    • Paused: Temporarily disabled but retains history
    • Draft: In progress, not yet protecting
  • Question being addressed
  • Expert-provided answer
  • Date created
  • Date last edited
  • Last edited by

2. Impact Metrics​

Codex automatically tracks the effectiveness and reach of each remediation:

  • Prevention Metrics:
    • Number of times this remediation prevented failures (highest usage)
    • Which past failure cases are now covered (broadest coverage)
    • Real-time updates as new cases are protected
  • Coverage Analysis:
    • Which past failures would have been prevented
    • Similar failure patterns identified
    • Automatic matching to relevant cases

3. Attribution System​

Codex provides unprecedented visibility into how each remediation improves your AI application:

  • Direct Impact Tracking:
    • Detailed logs of every prevention event where the remediation was triggered
  • Historical Coverage:
    • Links to all past failure logs this remediation now protects against

Important Notes​

  • Only one Active remediation and one Draft remediation are allowed per identical query
  • Paused remediations retain their historical impact data but won’t be used for new queries
  • Each log can only be “covered” by one Active remediation or one Draft remediation
  • All impact metrics update in real-time as new logs are created

The Remediations interface provides clear attribution for SME work, helping teams understand and measure the direct improvement to their AI application’s safety and reliability.