What Workable AI does
Workable AI combines an applicant tracking system with AI-assisted job briefs, sourcing, candidate evaluation, outreach, chat, and configurable recruiting-agent workflows.
Workable places agentic tasks inside a complete ATS rather than sending recruiters to a separate assistant. Requisitions, job-board distribution, sourcing, pipelines, interviews, offers, analytics, and governance provide the record the Agent reads before acting. Recruiters can configure sourcing parameters, screening thresholds, engagement rules, and scoring weights per account, job, or candidate, then stop the Agent globally or intervene with a candidate.
Pricing is comparatively visible but still variable. On the page reviewed July 15, 2026, Standard starts at $299 per month for organizations with 1–20 employees, while Premier and Enterprise are higher and company-size selectors change totals. Paid accounts receive 3,000 Agent credits; evaluation uses one credit, sourcing two, and candidate chat ten, while extra bundles start at $600 for 5,000 credits. Credits expire after one year and are non-refundable, and checkout or an order form is authoritative.
Workable says protected characteristics are excluded, low-confidence inferences pass to a person, each score includes reasoning, and actions are logged and reversible. Those are meaningful controls, not proof of unbiased outcomes. Recruiters must inspect whether job criteria or profile data act as proxies, test selection rates, disclose automated processing, honor deletion and accommodation requests, and prevent autonomous final hiring decisions. The platform reports ISO 27001, ISO 27017, SOC 2 Type II, and SOC 3; buyers should confirm plan and product scope.
How Workable AI works
A recruiter defines the role and ideal candidate profile inside Workable. The Agent searches profiles, evaluates candidates against configured criteria, sends approved outreach, chats to collect basic facts, scores responses and records, and can advance qualified people according to settings; actions inherit platform permissions and remain logged, reversible, and reviewable.
Translate the requisition into criteria
Recruiters define the role and ideal candidate profile, then review AI-assisted job language, screening questions, sourcing boundaries, weights, and Agent permissions. Remove unnecessary credentials and proxy variables before any candidate is evaluated.
Source and evaluate candidate records
The Agent searches available profiles and compares candidates with configured job criteria, consuming credits for sourcing or evaluation. Explanations help review, but missing data, stale profiles, and correlated attributes can still produce biased or weak recommendations.
Communicate and progress within guardrails
Approved outreach and chat gather basic facts, after which settings may score responses, shortlist candidates, or invite screening. Actions inherit ATS access, are logged and reversible, and should pause for human review, opt-outs, accommodations, ambiguity, or consequential decisions.
Review outcomes, controls, and credit use
Recruiters inspect messages, score reasoning, overrides, progression, group selection patterns, complaints, and credit consumption. Owners can adjust settings or use the kill switch, and should revalidate criteria, privacy notices, retention, and integrations before expanding autonomy.
How to set up Workable AI
Configure the hiring model
Define job-related criteria, permitted Agent actions, thresholds, candidate notice and consent, accommodations, appeal routes, and the recruiters accountable for every stage.
Set up the ATS safely
Create roles and permissions, connect required job boards and systems, configure retention and deletion, and import only accurate, authorized candidate and company data.
Build the job brief
Review generated job language, remove unnecessary requirements and proxies, set sourcing boundaries and scoring weights, and prepare approved outreach and screening questions.
Test Agent behavior
Use diverse representative profiles and ambiguous responses to inspect explanations, credits, opt-outs, messages, auto-advance, reversibility, integrations, and human handoffs.
Release and audit
Begin with recommendations and draft outreach, keep final decisions with people, review selection outcomes and overrides, monitor credit use, and recalibrate criteria when evidence changes.
Workable AI FAQs
How much does Workable cost?
Standard currently starts at $299 per month for 1–20 employees. Organization size, tier, billing cadence, add-ons, and Agent credits change the total.
How do Workable Agent credits work?
Evaluation uses one credit, sourcing two, and candidate chat ten. Paid accounts include 3,000 credits; the reviewed page lists 5,000 more for $600.
Can recruiters override the Agent?
Yes. Workable describes per-job and per-candidate controls, logged and reversible actions, an account-wide kill switch, and automatic stopping when a person engages.
How does Workable address recruiting bias?
Workable says it excludes protected characteristics and explains scores. Employers must still audit criteria and outcomes, investigate proxies, provide accommodations, and use human review.
Does the free trial include Workable Agent?
No. The pricing page says the 15-day trial covers Standard features but does not include Workable Agent.
Listing reviewed 2026-07-15. Product details and pricing can change; verify important terms on the provider's website.
Related Sales AI tools
Related AI guides
Reviews
Tell the community what you made, what worked, and what you wish you knew before starting.