AI Recruitment Screening Agent

An automated recruitment workflow that monitors job applications, downloads and analyzes CVs and essays using OpenAI assistants, then outputs structured screening results to help streamline the candidate evaluation process.

Purpose

No business context provided yet — add a context.md to enrich this documentation.

This workflow appears to be designed for automated recruitment screening, specifically for what seems to be a YET (Youth Entrepreneurship Training) program application process. It automatically processes new job applications by analyzing uploaded CVs and essays, providing AI-powered screening scores and summaries to help recruitment teams quickly identify qualified candidates.

How It Works

1. Application Monitoring: The workflow watches a Google Sheets document for new job applications
2. Latest Entry Selection: When new rows are added, it identifies the most recent application based on timestamp
3. CV Processing: Extracts the CV file ID from Google Drive links and downloads the document
4. CV Analysis: Uploads the CV to OpenAI and uses a specialized "file analyzer" assistant to extract text and analyze experience, education, and skills
5. Essay Processing: Similarly processes the entrepreneurship essay submission from applicants
6. Essay Scoring: Uses an "Essay analyzer" assistant to score the essay based on specific criteria (0-100 points)
7. Data Structuring: Converts AI analysis results into structured JSON format
8. Results Output: Writes the original application data along with AI analysis to a separate results spreadsheet
9. Cleanup: Deletes uploaded files from OpenAI to manage storage

Workflow Diagram

graph TD
    A[Google Sheets Trigger - CV] --> B[Code3 - Get Latest Row]
    A2[Google Sheets Trigger - Essay] --> B2[Code4 - Get Latest Row]

    B --> C[Code - Extract CV File ID]
    B --> K[Google Sheets - Output Results]

    C --> D[HTTP Request - Download CV]
    D --> E[OpenAI - Upload CV File]
    E --> F[OpenAI6 - Update CV Assistant]
    F --> G[OpenAI2 - Analyze CV]
    G --> H[Code2 - Structure Output]
    H --> I[OpenAI8 - Delete CV File]

    B2 --> C2[Code1 - Extract Essay File ID]
    C2 --> D2[HTTP Request1 - Download Essay]
    D2 --> E2[OpenAI3 - Upload Essay File]
    E2 --> F2[OpenAI9 - Update Essay Assistant]
    F2 --> G2[OpenAI5 - Score Essay]
    G2 --> I2[OpenAI10 - Delete Essay File]

Trigger

Google Sheets Trigger (2 instances) - Monitors: "Copy of SCREENING WORKING SPACE" spreadsheet, "YET_applications" sheet - Event: Row added - Frequency: Every minute polling - Triggers when new job applications are submitted to the form

Nodes Used

Node Type	Purpose	Count
Google Sheets Trigger	Monitor for new applications	2
Code	Extract file IDs, get latest entries, structure data	4
HTTP Request	Download CV and essay files from Google Drive	2
OpenAI (File Upload)	Upload documents to OpenAI for analysis	2
OpenAI (Assistant Update)	Attach files to specialized assistants	2
OpenAI (Assistant Query)	Analyze CV and score essay	2
OpenAI (File Delete)	Clean up uploaded files	2
Google Sheets	Output screening results	1
Sticky Note	Documentation and workflow organization	9

External Services & Credentials Required

Google Services

• Google Sheets Trigger OAuth2: For monitoring application submissions
• Google Service Account: For writing results to output spreadsheet
• Google Drive: File storage for CV and essay uploads (accessed via direct download URLs)

OpenAI

• OpenAI API: For file processing and AI analysis
• Pre-configured Assistants:
  • asst_XE7zzsVBnrgP6DIhsldk3LPh - "file analyzer" for CV processing
  • asst_BX1yz8XVvktnWJitik7Q0NCg - "Essay analyzer" for essay scoring

Environment Variables

No explicit environment variables are used. All configuration is handled through: - Google Sheets document IDs (hardcoded in nodes) - OpenAI assistant IDs (hardcoded in nodes) - Credential references stored in n8n

Data Flow

Input

• Source: Google Sheets form responses containing:
  • Applicant personal information (name, email, phone, region, etc.)
  • CV file (Google Drive link)
  • Essay submission (Google Drive link)
  • Education and experience details

Processing

• CV Analysis: Extracts text via OCR, summarizes experience and education, provides overall rating
• Essay Scoring: Evaluates entrepreneurship essay against specific criteria (0-100 point scale)

Output

• Destination: "AI Recruitment" Google Sheets document
• Data: Original application data plus AI analysis results
• Format: Structured rows with applicant info and screening scores

Error Handling

• HTTP Request1: Configured with "Continue on Error" for essay download failures
• OpenAI8 & OpenAI10: File deletion operations set to continue on error to prevent workflow failure
• Retry Logic: CV and essay analysis nodes have retry enabled for API reliability

Known Limitations

Based on the workflow structure: - Processes one application at a time (no batch processing) - Requires specific Google Drive link format for file extraction - Dependent on OpenAI assistant availability and rate limits - Essay scoring is specific to entrepreneurship topic with "salubrious" word requirement - No validation for file types or sizes before processing

Related Workflows

No related workflows specified in the current context.

Setup Instructions

1. Import Workflow: Import the JSON into your n8n instance

2. Configure Google Credentials:

   • Set up Google Sheets Trigger OAuth2 for monitoring applications
   • Configure Google Service Account for writing results

3. Configure OpenAI:

   • Add OpenAI API credentials
   • Create or configure the required assistants:
     • CV analyzer assistant with appropriate prompts
     • Essay analyzer assistant with scoring criteria

4. Update Spreadsheet References:

   • Replace hardcoded Google Sheets document IDs with your actual sheets
   • Ensure proper column mapping in the output node

5. Test the Workflow:

   • Add a test application to your source spreadsheet
   • Verify CV and essay files can be downloaded
   • Check that AI analysis completes successfully
   • Confirm results are written to output sheet

6. Activate: Enable the workflow to begin monitoring for new applications

Note: This workflow is currently marked as inactive and may require additional configuration before production use.