🕵️‍♂️ Fake Applicant Detector (FAD)

An intelligent, agentic system designed to identify fraudulent "Ringer" candidates, AI-generated resumes, and ghost LinkedIn profiles. By cross-referencing live LinkedIn and Github data against uploaded resumes, the agent calculates a TrustScore based on real-world behavioral signals.

🚀 How it works

The Fake Applicant Detector orchestrates a multi-step investigation loop:

1. Resume Extraction: Parses text from Google Drive, Dropbox, or public PDF links.
2. Identity Sync: Cross-references LinkedIn dates and titles with the Resume to find discrepancies (>3 months).
3. GitHub Deep Dive: Checks account age and activity bursts. Flags "Senior" developers with brand-new GitHub accounts.
4. Company Audit: Investigates the legitimacy of current/past employers. If a company looks like a "shell," the agent audits its domain age and digital footprint.
5. AI Slop Check: Analyzes text for high-entropy AI phrasing common in generated resumes.

📥 Input

The Actor requires a LinkedIn URL, a Resume link, and an AI provider choice.

Field	Type	Description
linkedinUrl	String	Public LinkedIn profile link of the candidate.
resumeUrl	String	URL to the resume (Direct PDF, Google Drive, etc.).
modelProvider	Enum	Choose between google (Gemini 2.0) or openai (GPT-4o).

Example Input:

{
    "linkedinUrl": "https://www.linkedin.com/in/arjun-nambiar-439aaa255",
    "resumeUrl": "https://drive.google.com/file/d/1wX_LY.../view",
    "modelProvider": "google"
}

📤 Output

The results are saved to the default Apify Dataset. Each item contains a detailed markdown analysis and a structured verdict.

Example Output Data:

{
  "TrustScore": 75,
  "Verdict": "SUSPICIOUS",
  "RedFlags": [
    "Current company 'Kanvarshi Motors' not found on LinkedIn",
    "Employment date discrepancy in experience duration",
    "GitHub account created 2 months ago despite 4 years of claimed experience"
  ],
  "WritingStyle": "Human",
  "SocialFootprint": {
    "linkedInFound": true,
    "githubFound": true,
    "portfolioFound": false
  },
  "Analysis": "### Breakdown of Findings...\n\n1. **Identity Sync**: Resume lists Kanvarshi Motors which does not exist on LinkedIn...\n2. **GitHub Audit**: Profile is too new for a 'Senior' level claim...",
  "candidateUrl": "https://www.linkedin.com/in/arjun-nambiar-439aaa255",
  "status": "COMPLETED",
  "timestamp": "2026-01-13T13:55:05.672Z"
}

📋 Features

• Behavioral Footprint: Identifies GitHub accounts with suspicious activity bursts.
• Live Link Verification: Automatically visits portfolio/project links to ensure they aren't dead or generic templates.
• Account Age Analysis: Flags brand-new social profiles used for job application fraud.
• TrustScore Calculation: A logical 0-100 score based on evidence found across 4 different data sources.

⚠️ Disclaimer

This tool is intended for recruitment assistance and fraud screening only. It uses AI to interpret public data; always perform manual verification and interview-based vetting before making final hiring decisions.