🛡️ HireShield - AI Candidate Verification AI-powered candidate verification. Paste resume → get instant risk assessment with explainable scoring. ## 🔍 What It Does Paste resume → Get comprehensive verification assessment No required fields. Just paste the resume text and all candidate info is auto-extracted. ## 📊 Signal Categories | Category | Checks | |----------|--------| | Identity | Name/email alignment, location consistency | | Timeline | Employment overlaps, gaps (only if dates provided) | | Company | Domain verification, Wikipedia, DNS checks | | GitHub | Contribution depth, fork detection (role-weighted) | | Fingerprint | Resume duplicate detection, cluster patterns | | Communication | Disposable email, MX records, VoIP detection | ## 🎯 Key Features - AI Resume Parsing - Gemini extracts structured data (optional) - Resume Fingerprinting - Detects duplicate/template reuse across candidates - Cluster Detection - Flags coordinated submission patterns - Role-Based Weighting - Designers not penalized for missing GitHub - GDPR Compliant - 180-day data retention, hashed emails ## ⚠️ Verification Levels | Score | Level | Action | |-------|-------|--------| | 0-30 | LOW | Proceed with standard process | | 31-60 | MEDIUM | Additional verification recommended | | 61-80 | HIGH | Enhanced verification required | | 81-100 | CRITICAL | Proceed only after enhanced verification | ## 📥 Inputs | Field | Required | Description | |-------|----------|-------------| | resumeText | ✅ | Paste resume - auto-extracts everything | | roleType | | fullstack, backend, frontend, designer, pm, other | | candidateType | | full-time, student, freelancer, contractor | | previousRiskScore | | For risk trajectory tracking (0-100) | ## 📤 Output Schema json { "version": "3.0.0", "candidate": { "name": "John Smith", "email": "john@company.com", "targetRole": "backend" }, "risk_assessment": { "overall_score": 15, "risk_level": "LOW", "risk_trend": "improving", "action": "Proceed with standard process", "summary": "Profile verified with minimal concerns..." }, "signals": { "negative": [...], "positive": [...], "total_negative": 1, "total_positive": 2 }, "fingerprint": { "semantic_hash": "abc123", "structural": {...} } } ## 🔧 Environment Variables | Variable | Default | Description | |----------|---------|-------------| | GEMINI_API_KEY | - | Enables AI parsing and explainability | | TENANT_ID | default | Multi-tenant fingerprint isolation | | DATA_RETENTION_DAYS | 180 | GDPR retention period | | SIMILARITY_THRESHOLD | 0.85 | Fingerprint duplicate threshold | ## 🚀 Quick Start json { "resumeText": "John Smith\nSan Francisco • john@company.com\nSenior Engineer at Google (2020-2024)...", "roleType": "backend" } ## ✅ Enterprise Compliance - No scraping - Consent-based data only - Explainable - Clear verification reasoning - GDPR - Data retention and email hashing - Neutral language - "Verification concern" not "risk" ## 🔐 Tenant Isolation (Important!) Each company gets isolated fingerprint data. Set your tenantId in the input to prevent cross-company data mixing. json { "resumeText": "...", "tenantId": "your-company-name" } | Company | tenantId Input | |---------|----------------| | Your Company | "tenantId": "your-company" | | Another Company | "tenantId": "another-company" | > ⚠️ Warning: If you don't set tenantId, you share fingerprint data with all other default users! ## 💰 Pricing $0.02 per candidate analysis Includes AI parsing, fingerprint matching, 6 verification categories, and explainable assessment. ## 📄 License ISC