A platform team wants repeatable ML infrastructure in multiple environments without manual console drift. Which two practices are appropriate? Select 2 answers.

1. Use the resource display name as a complete version history.
2. Embed all permanent administrator credentials in the template.
3. Validate infrastructure changes and permissions in a controlled deployment pipeline.
4. Define infrastructure as versioned code with environment-specific configuration.
5. Allow untracked manual edits as the only source of production truth.