Build Your AI-Powered Hiring Pipeline in 5 Steps
From configuring job boards to deploying bias-aware AI screening in under 15 minutes.
5-Step Hiring Setup
Create Your First Job
Navigate to Recruit > Jobs > Create. Enter title, department, location, and requirements. Use the AI Job Architect (optional): describe the role in plain English and the AI generates a market-aware job description optimized for candidate conversion. Set salary range and employment type.
Configure AI Screening
Navigate to Recruit > Settings > AI Screening. Enable resume parsing, skill profile extraction, and job-match scoring. Set bias detection thresholds (default: flag all severity levels). Choose your preferred AI provider (a high-capability model is recommended for best parsing accuracy). Set the confidence threshold for auto-surfacing top candidates (default: 85+).
Set Up Video Interviews (Optional)
Navigate to Recruit > Video Templates > Create. Add questions with thinking time and recording time limits. Configure retake settings and expiry. Set welcome/closing messages. Send template to candidates via email with unique presigned link.
Configure Job Board Syndication
Navigate to Recruit > Syndication. Connect the 4 supported boards (LinkedIn, Indeed, Glassdoor, ZipRecruiter) via their partner API keys. Until a board's key is configured it runs in a clearly-flagged test mode (no silent fake posts). Select boards to publish to. Configure board-specific formatting and tracking URLs. Set auto-repost schedule if desired.
Enable GDPR Compliance
Navigate to Recruit > Settings > GDPR. Set retention period (default: 2 years). Enable automated weekly purge. Configure consent requirements (data processing, background check, video interview). Test right-to-be-forgotten flow with a test candidate.
typescript
// Example: List open jobs through the recruit MCP toolconst jobs = await aeion.recruit({ action: "list_jobs", filters: { status: "open" }, limit: 20,});
// Example: Score a candidate against a job (AI screening + 5-vector bias check)const result = await aeion.recruit({action: "score_candidate",id: "cand_abc123",jobId: "job_xyz789",});// result.overallScore === 87 (strong_match)// result.biasReport.flags === [] (no bias detected)// result.matchAnalysis.matchedSkills === ["React", "TypeScript", "Node.js"]
// Example: Extract skills from a candidate's resumeconst skills = await aeion.recruit({action: "extract_skills",id: "cand_abc123",});
// Example: Detect bias in a job description before postingconst bias = await aeion.recruit({action: "detect_bias",id: "job_xyz789",});
// Example: Schedule an interviewawait aeion.recruit({action: "schedule_interview",applicationId: "app_123",interviewerIds: ["user_456"],scheduledAt: "2026-06-01T15:00:00Z",duration: 60,});
// Example: Create an offerconst offer = await aeion.recruit({action: "create_offer",candidateId: "cand_abc123",jobId: "job_xyz789",terms: {jobTitle: "Senior Software Engineer",department: "Engineering",employmentType: "full_time",startDate: "2026-07-01",salary: 180000,salaryCurrency: "USD",salaryPeriod: "annual",expiresAt: "2026-06-15",},});
// GDPR data exports are also available directly from a candidate's profile// in the admin UI, or through the platform API for automated compliance workflows.Each vector runs its own dedicated check: Name (ethnic/gender patterns), Location (socioeconomic signals tied to geography), Education (prestige comparison), Career Gap (parental leave, military, caregiving detection), and Age (implicit age signals such as graduation-year math). If detected, the severity (low/medium/high) and a guidance note ("Focus on qualifications, not name") are surfaced to the recruiter. High-severity flags trigger ethics review workflows.
Candidates upload video directly via a secure, time-limited link — no server proxy needed. Playback uses a separate secure link with its own expiration. Audio is transcribed automatically and scored for sentiment (positive/neutral/negative).
When you mark a candidate as hired, Aeion Recruit and Aeion HR automatically exchange the record — no manual export or sync step. Candidate data (name, email, location) maps directly to employee fields, which immediately triggers payroll setup and IT provisioning workflows. Zero manual data entry.
Yes. In Recruit > Settings > AI Screening, you can set severity thresholds for each bias vector. By default, all severity levels are flagged. You can set to "medium and high only" or "high only" to reduce noise. The guidance note is always shown regardless of threshold.
Syndication tracks each posting by job and publish date to detect duplicates. When republishing, it checks whether a listing already exists on each board and updates it rather than creating a new one. Board-specific listing IDs are stored so updates land on the right posting.
An automated weekly check scans for candidates who've passed their retention window (default: 2 years / 730 days). Expired candidates are hard-deleted, including related applications, interviews, and scorecards. Anonymous aggregate counts (funnel analytics) are preserved. Before deletion, you can generate a data export (JSON/PDF) for compliance records.
Matching analyzes three dimensions: Skill Match (technical + soft skills vs. requirements), Experience Match (years, relevance, industry overlap), and Culture Fit (work style alignment). Each dimension scores 0-100. The overall score is a weighted average. Recommendations range from "strong match" (85-100) to "no match" (0-24).
Yes. The AI engine is configurable in Platform > AI Settings, and any registered provider (OpenAI, Anthropic, Google AI, Groq) works. If AI is temporarily unavailable, screening falls back to rule-based keyword matching so hiring never stalls.