How Mixpanel uses Greptile
Mixpanel uses Greptile to automatically review pull requests, most of which are written by coding agents such as Claude Code and its in-house agent TAL. Greptile gives codebase-wide feedback before a human reviewer looks at a PR, and also reviews TAL's own output. Greptile reports about 9,300 code reviews per month at Mixpanel.
View captured proofThe use case: how Mixpanel deploys Greptile
Mixpanel uses Greptile to automatically review pull requests, most of which are written by coding agents such as Claude Code and its in-house agent TAL. Greptile gives codebase-wide feedback before a human reviewer looks at a PR, and also reviews TAL's own output.
More reported results
Figures published in Greptile’s case study. They are not independently measured by usedby.
~4,800+ Issues surfaced per month
Greptile reviews ~2,900 PRs a month
posts ~2,400 P0–P2 findings inline
In the customer’s words
The thought of adding an agent to PRs to do code reviews, a potential slop cannon, was extremely contentious. We took the leap and we're so glad we did.Other AI tools used by Mixpanel
Other public relationships recorded on usedby.
Explore Mixpanel’s AI stackGreptile users in Data & Analytics
Companies in the same sector, with public evidence.
See all Greptile customersGo straight to the source
Based on Greptile’s published case study. We check the public evidence; the results are reported by the vendor.
greptile.com- Last checked
- Nov 7, 2026
Open source capture What the story claims, and what we checked
What we compared with the page.
- The figure: ~9,300Checked
Printed word for word on the page, near the name of Mixpanel.
- The passage quoted aboveChecked
Copied word for word from the page, near the name of Mixpanel.
- The numbers in our summaryChecked
Each one is printed on the page.
- Mixpanel uses GreptileChecked
Confirmed line.
- The result itselfNot checked
We quote it; we did not measure it.
Greptile reviews ~2,900 PRs a month and posts ~2,400 P0–P2 findings inline. It runs on TAL's output too. That costs less than putting a review step inside TAL and returns faster in wall-clock time.








