Community stats
Aggregate data from resumasher installs that opted into the community tier. Every number on this page is computed from the underlying telemetry database via a SECURITY DEFINER Postgres function; no raw rows, no names, no resume content, no company details leave the server. Updated live every time you refresh.
If you opted in and want your data erased: run resumasher
telemetry delete and the right-to-erasure endpoint will wipe every
event tied to your installation ID.
Loading live data…
Runs per day
AI CLI host
Which CLI students are running resumasher in.
Model
Top 15 models self-reported by the orchestrator. Gemini in auto-routing mode shows up as different variants depending on load.
Fit score distribution
How well the fit-analyst scored candidate-to-JD alignment (0 = bad, 10 = strong).
Seniority of targeted roles
Bucketed via server-side CASE WHEN from raw LLM output. "Early-Career/Graduate" from weaker models gets normalized to junior here.
Placeholder fill choices
When the tailor emits an [INSERT…] placeholder, the student picks one of three: give specifics, soften to a no-metric alternative, or drop the bullet.
Failures by phase
Which pipeline phase most frequently hard-stops a run, grouped by error class. Phase 0 = setup, 1 = intake, 2 = mine, 3 = fit, 4 = company research, 5 = tailor, 6 = cover letter + prep, 7 = placeholder fill, 8 = PDF render, 9 = summary.
Data generated at —.
Could not load stats. Retry.