Rules that route and notify — with people in the loop.
Hiring Policy is an event-driven rule engine for recruiters. Define if-this-then-that rules that act on candidates automatically as assessment milestones occur — routing them forward and sending notifications — while the decisions that affect a candidate stay strictly with a person.
Hiring Policy · rule
When
An assessment is graded
Event-driven — the rule reacts as candidate events occur.
If
The confirmed score meets your bar
Conditions scope the rule to the candidates you mean.
Then — in priority order
Route to an assessment
Send the candidate the next assessment in the funnel.
Route to a reviewer
Assign the candidate to a specific team member.
Send an automated email
Notify the candidate or the team with a templated message.
Consequential actions are held for a human. Anything that would affect a candidate’s outcome is suppressed until a person records a decision — it never auto-executes.
Every match and action is written to an execution log — what fired, when, and what it did.
Rule Engine Structure
Anatomy of a rule
A hiring-policy rule is a small, readable piece of automation. You decide what it applies to, what starts it, what must be true, and what it does — in that exact order.
Scope
Which jobs, pipelines, or candidate pools the rule applies to. Avoid global side-effects by targeting precise requisitions.
role == "Senior ML"Trigger
The event that initiates rule evaluation — such as a completed assessment, verified identity telemetry, or submission deadline.
event.graded_at != nullConditions
The empirical checks that must hold before execution. Combines normalized score percentiles, skill matches, and test validity metrics.
score >= 78 && integrity == 100Actions
What happens next, run sequentially in the prioritized chain you designate — routing, reviewer queues, and templated emails.
dispatch(stage_2, notify_lead)Action Capabilities
What a rule can do today
Three primary actions are live today — funnel routing and automated communications. Each runs automatically the millisecond a rule matches, in the explicit priority order you set.
Route to an assessment
Automatically send the candidate the next assessment in your evaluation funnel the moment the trigger fires. No waiting for manual recruiter queue sweeps.
Route to a reviewer
Assign the candidate directly to a specific team member, domain peer, or rotating hiring committee panel so the right expert picks up the deep qualitative review immediately.
Send an automated email
Dispatch a templated, branded message to the candidate or internal hiring team — a status nudge, next-step preparation guide, schedule request, or transparent progress confirmation.
Ethical & Safe Architecture
Built for human oversight over automated decisions
Automation is only safe if consequential outcomes stay firmly with a person. Hiring Policy is architected from the bare metal around that conviction — routing and notifications execute smoothly on their own, but any decision affecting a candidate’s status is permanently held for an affirmative human judgment.
Consequential actions are held
Candidate-affecting actions (rejection, hiring offer, salary qualification) are physically suppressed in code until a recruiter or hiring manager records a rationale. They can never auto-execute.
Fully logged execution trace
Every rule trigger match, condition outcome, and dispatched webhook is serialized to an append-only ledger. When compliance officers inspect outcomes, every step is mathematically reviewable.
Hardened for absolute reliability
Actions are strictly idempotent and claimed atomically in transaction before runtime execution. Network retries, timeout recoveries, or concurrency spikes cannot trigger a duplicate action.
Where we draw the line
We are unapologetically deliberate about what software should and should never do. Hiring Policy routes and notifies — it never decides who gets hired or rejected.
Explicitly Prohibited Capabilities (Engine Level)
Auto-hire, auto-reject, or auto-eliminate candidates
No algorithm or threshold is ever permitted to terminate a candidacy without affirmative human confirmation.
Silently mutate a candidate’s lifecycle status
Status changes require signed actor attribution in the applicant tracking audit log.
One-click bulk discard of held actions
We deliberately prevent high-volume “blind rubber-stamping” that circumvents human evaluation.
Escalation or penalty SLA timers on pending evaluations
Candidates are never timed out or dropped due to reviewer scheduling backlogs.
What it does do:
Run the routine, repetitive routing, invitations, and messaging instantly. Keep a human reviewer on every consequential fork, and log the entire chain immutably.
Assessments provide the signal — Hiring Policy acts, carefully.
Automate the busywork. Keep the judgment.
Bring a hiring funnel, and we will demonstrate how Hiring Policy eliminates days of candidate latency while safeguarding fairness and institutional review where it matters most.