What hiring analytics covers
get_pipeline_overview gives a top-level snapshot: totals by stage overall and per job for your busiest roles, flagged and tagged counts, and the most recent application date, the starting point for any broad how's-hiring-going question. list_applicants filters down to a specific set by job, stage, score range, or tag.
get_applicant_profile pulls a deep view of one candidate: scores, strengths and weaknesses, challenge results, and which screening rules matched them. compare_applicants puts 2 to 10 candidates side by side with a computed average, max, and min per score, ready to chart.
get_hiring_stats covers trends over a date range, defaulting to the last 90 days: applications and average score per week, stage conversion counts, and average scores by job or by coding challenge. analyze_hiring_funnel and get_dossier_data, from the related interview-intelligence tools, dig further into where a funnel narrows and pull a fuller candidate dossier when one profile isn't enough.
A walkthrough: figuring out where candidates are dropping off
An employer asks: "why does it feel like we're losing good candidates in screening?" The agent starts with get_pipeline_overview, sees a large screening stage relative to qualified, then calls get_hiring_stats for stage conversion counts over the last 90 days, confirming a real drop-off between screening and qualified rather than a one-week blip.
It calls analyze_hiring_funnel to check whether the drop concentrates on a specific job or score range, finding it's mostly candidates scoring 6 to 7 overall, a band with no screening rule currently covering them. The employer then asks it to compare three specific candidates sitting in that band right now.
compare_applicants returns all three side by side with their scores and top strengths and weaknesses, giving the employer something concrete to decide from rather than a guess.
Boundaries
What analytics can and can't touch
Every tool in this group is read-only; none of them can move a candidate, change a job, or write anything back to your pipeline, so asking a question here never risks an unintended action, and there's no approval card to review because nothing is being changed.
list_applicants and get_applicant_profile never return a candidate's email or phone number, even when pulling a deep profile with strengths, weaknesses, and challenge results attached. get_hiring_stats defaults to the last 90 days unless a different range or a specific job is given, and its output is described as chart-ready rather than a finished chart, since rendering it is the interface's job, not the tool's.
That distinction matters mostly for how fast a follow-up question can be answered: because nothing here writes anything, the agent can run five or six of these tools back to back inside one conversation without pausing for a confirmation at any point.
Frequently asked questions
Can the agent accidentally change my pipeline while answering a data question?
No. Every tool used for hiring analytics, from the pipeline overview to comparing candidates, is read-only. None of them can move a stage, tag a candidate, or write anything back.
What time range does hiring stats cover by default?
The last 90 days unless you ask for a different range or scope it to one specific job. It returns applications and average score per week, stage conversion counts, and average scores by job or coding challenge.
Can I compare specific candidates against each other?
Yes, compare_applicants puts 2 to 10 candidates side by side with their scores, top strengths and weaknesses, and a computed average, max, and min per score metric.
Does the agent return candidate contact information when I ask for analytics?
No. list_applicants and get_applicant_profile are both built to withhold email and phone even when returning a full profile.
Can the agent tell me where in the funnel candidates are actually dropping off, not just how many applied?
Yes, analyze_hiring_funnel looks specifically at where a funnel narrows, whether that's concentrated on one job or one score range, rather than just totaling applications by stage.
Related pages
Ask how hiring is actually going
Ask the agent for a pipeline overview, a funnel breakdown, or a side-by-side comparison of your current finalists.