What it does
A rule is a set of conditions checked against a completed interview, paired with an action. Conditions can reference any of the five AI scores, a technical challenge score, the candidate's experience level, salary estimates pulled from the interview, key skills mentioned, or the strengths and weaknesses the AI wrote up.
Fourteen operators are available for comparing those fields, and conditions can be nested with AND and OR logic up to five levels deep and twenty-five conditions total, so a rule can be as simple as one score threshold or as specific as a multi-part filter across skills, experience, and salary range at once.
How it works
- 1
Build a rule from a template or from scratch
Rule templates cover common patterns to start from, or a rule can be built condition by condition against the fields available for that job.
- 2
Test it against real data before it goes live
The live test-a-rule tool runs the rule against existing candidates and shows a match count, so a recruiter can see how many people it would actually catch before turning it on.
- 3
The interview finishes and the rule fires
As soon as a candidate's AI interview is scored, every active rule for that job evaluates automatically. No one has to trigger it manually or run a batch job at the end of the day.
- 4
The matched action executes immediately
A matching rule can tag the candidate, set their pipeline stage, set a priority level, flag them, or notify a teammate, chosen when the rule was built.
- 5
Match counts stay visible over time
Each rule keeps a running match count, so a recruiter can tell at a glance whether a rule is catching a handful of candidates or most of the applicant pool, and adjust the thresholds accordingly.
Fourteen operators cover more than pass or fail
Rules are not limited to greater-than checks on a single score. The operator set lets a rule check ranges, exclusions, and text matches across the evaluable fields, so a condition can be as precise as technical score above 7 and experience level at least mid and salary estimate under a set ceiling, combined in one rule.
Five actions, chosen per rule
Tag adds a label to the candidate. Set_stage moves them in the pipeline. Set_priority raises or lowers how urgently they surface. Flag marks them for attention. Notify sends an alert to a teammate. A single job can run several rules with different actions covering different score ranges at once.
Nesting handles the real-world exceptions
Twenty-five conditions across five nested levels of AND and OR means a rule can express something like a strong technical score OR a strong behavioral score, AND acceptable salary expectations, AND not already tagged rejected, in one rule instead of three separate ones a recruiter has to remember to check in order.
Who it's for
Who this is for
This is built for roles with enough applicant volume that manual triage is the bottleneck, a high-volume support role, a seasonal hiring push, or a technical role with a wide funnel at the top. A single low-volume role can run without any rules at all and rely on manual review instead.
Frequently asked questions
What data can a screening rule check?
All five AI interview scores, the technical challenge score where one exists, experience level, salary estimates from the interview, key skills, and the strengths and weaknesses the AI wrote in its assessment. Any of these can be combined into a rule's conditions.
What happens when a rule matches a candidate?
The rule's chosen action fires immediately: tag, set_stage, set_priority, flag, or notify a teammate. This happens automatically as soon as the interview is scored, without a recruiter opening the pipeline.
Can I test a rule before turning it on?
Yes. The live test-a-rule tool runs the rule against your existing candidates and shows how many it would match, so you can check a rule's reach before it starts acting on new interviews.
How complex can one rule get?
Up to 25 conditions nested across 5 levels of AND and OR logic, using 14 comparison operators. Most rules are simpler than that ceiling, but it is there for roles with multiple hard requirements.
Related pages
Let scoring do the first pass on every applicant
Set up a rule against your first job's scoring bar and watch it tag, prioritize, or flag candidates the moment their interview finishes.