An agent is a specialist. It runs one part of hiring, automatically.

An agent is triggered by a defined event in your hiring workflow — a candidate being shortlisted, an interview being scheduled, a document being uploaded, an offer being accepted, or a time interval being reached. Agents do not run continuously; they run precisely when the defined event occurs.

Conditions filter who an agent acts on. You configure the agent to run only on candidates who match specific criteria — role type, skill set, location, experience level, or any field in your ATS. This prevents agents from acting on candidates they are not relevant to.

Every agent ends with a handoff. The Voice Agent delivers a scorecard and transcript. Interview Copilot surfaces questions and candidate context in Teams. The Document Validation Agent flags mismatches for review. The output is always structured and actionable — not raw data for the recruiter to interpret.

Yes. No agent makes a final decision. Recruiters can review any agent output, override any flag, and advance or reject candidates regardless of the agent's score. Every override is logged with a timestamp and the user's identity.

Every agent action is logged in the job and candidate collaboration tabs. Workflow history is available for every agent, showing what triggered it, which candidates it acted on, what action it took, and what the outcome was.

Yes. An agent's output can feed the next stage. For example, a Voice Agent completes a pre-screening call and delivers a scorecard. A recruiter advances the candidate. Interview Copilot then auto-triggers when the interview is scheduled, reading the prior round's context. Each agent knows the full candidate story up to that point.

Pre-built agents like the Voice Agent, Interview Copilot, and Document Validation Agent are configured in Admin Settings and can be active within minutes of enablement. Custom agents built in the Agent Builder — with your own trigger, condition, and action — can typically be designed, tested, and deployed in a single session.

Yes. Conditions let you configure agents to run only on specific role types, departments, locations, or job functions. The same agent setup can behave differently for volume hiring and leadership hiring by configuring separate condition sets.

Yes. Agent Builder and all five pre-built agent categories are included in the platform at no additional cost. AI-heavy compute tasks such as voice screening, document AI, and fraud detection are billed at cost as pass-through.

Building a custom integration requires engineering time, ongoing maintenance, and a dependency on API stability. Agent Builder runs inside RippleHire, uses hiring-native data, and requires no engineering resource to configure, test, or update. When models drift or job boards change, RippleHire manages the maintenance — your team manages the workflow logic.