- Public, detailed case materials
- The vendor publishes customer narratives and operational descriptions (Anyshift, PostHog, Metabase), but fully detailed, independently verified case studies with methodology, baselines, and ROI calculations were not located in public sources.
- Vendor-authored PostHog operational narrative
- Mendral published a blog post describing PostHog CI volume for a specific week with vendor-provided metrics such as:
- a very large number of CI jobs across many workflow runs
- a very large volume of log lines
- a very large number of test executions
- Critical evaluation
- These numbers are vendor-published and not independently audited. They are useful as an illustrative scale archetype for teams considering such tooling, but buyers should validate applicability to their own environments.
- Public PRs as operational evidence
- The vendor indicates there are many merged PRs authored by the agent in public repositories (for example, PostHog). This provides a way for buyers to inspect PR quality, scope, and acceptance patterns.
- Critical evaluation
- PR volume alone does not indicate net value; it could also reflect increased review workload. Buyers should sample PRs to assess correctness, code quality, and alignment with internal standards.
- Vendor metrics without independent validation
- Vendor-published metrics (median diagnosis time, investigations per month, log compression, query scan sizes) do not include detailed methodology or definitions in publicly available materials.