Anonymized PostgreSQL queries against the shipped schema/views. Aggregate only — no profile or company identifiers in the result set. Apply schema.sql first.
Use lifecycle_duration_observable from v_jobs_current before treating lifespans as complete (only prospective rows have a trustworthy open clock).
SELECT
percentile_cont(0.5) WITHIN GROUP (
ORDER BY EXTRACT(EPOCH FROM (inactive_detected_at - first_seen_at)) / 86400.0
) AS median_days_to_inactive,
count(*) AS n
FROM v_jobs_current
WHERE lifecycle_duration_observable
AND inactive_detected_at IS NOT NULL
AND status IN ('inactive_inferred', 'offline_confirmed', 'deleted_confirmed');SELECT
count(*) FILTER (WHERE reopened_count > 0) AS reopened,
count(*) AS ever_inactive,
round(
100.0 * count(*) FILTER (WHERE reopened_count > 0) / nullif(count(*), 0),
1
) AS reopen_pct
FROM v_jobs_current
WHERE inactive_detected_at IS NOT NULL
OR reopened_count > 0;SELECT
percentile_cont(0.5) WITHIN GROUP (ORDER BY views_per_hour) AS median_views_per_hour,
percentile_cont(0.5) WITHIN GROUP (ORDER BY apps_per_hour) AS median_apps_per_hour,
count(*) AS n
FROM v_job_dynamics
WHERE is_short_window IS TRUE
AND views_per_hour IS NOT NULL;SELECT status, count(*) AS n
FROM v_jobs_current
GROUP BY status
ORDER BY n DESC;SELECT
status,
count(*) AS n,
percentile_cont(0.5) WITHIN GROUP (
ORDER BY EXTRACT(EPOCH FROM (now() - first_seen_at)) / 86400.0
) AS median_age_days
FROM v_candidates_current
GROUP BY status
ORDER BY n DESC;Candidate listing discovery has been confirmation-only since August 2026; these aggregates still mature on the stored cohort.