A system based on GPT-4o and n8n that analyzes resumes, contacts candidates and CTO in complete autonomy
Written by Sergio Sentinelli · CTO AIONSoft · January 10, 2025 · 8 min read
In the HR world, resume management is often a long and repetitive process. Manual screening, skills assessment, sending standardized emails: operations that take precious time away from recruiters for more strategic activities.
With this experiment I wanted to demonstrate how a system based on Large Language Models and n8n automations can manage the entire selection process autonomously: from CV analysis to sending communications, through comparison with open positions.
A virtual HR that works 24/7, analyzes hard and soft skills, compares with company job openings and responds to the candidate with programmed empathy.
The "CV Parser" workflow was built entirely in n8n, the open-source automation platform, and leverages GPT-4o as its analytical brain. Here's the complete journey each resume goes through:
The candidate fills out an n8n form by uploading their CV in PDF format. The trigger automatically activates the flow.
The Extract from File node extracts all text from the PDF document, transforming it into input processable by AI.
The AI Agent analyzes the text and generates a structured JSON containing: technical skills, work experiences, soft skills, education, languages and certifications.
The system automatically sends a first receipt confirmation email to the candidate, personalized with their name and thanking them for their interest.
A second email is sent to the CTO (Sergio Sentinelli) containing the analyzed CV, structured JSON and preliminary profile assessment.
The system queries a Google Sheet with active job openings and calculates the percentage match between candidate skills and sought requirements.
The complete flow from CV upload to final response
Screenshot of the actual n8n workflow used for HR automationBeyond the technical result, this project demonstrates 4 concrete advantages of AI automation applied to HR processes:
What required 15-20 minutes of manual work per CV now happens in less than 30 seconds. The recruiter can focus on interviews and human evaluation.
Each CV is analyzed with the same objective criteria. No risk of bias from fatigue, haste or undocumented subjective preferences.
The system can handle 10 applications per day as well as 1000, without performance degradation. Automatic scalability without additional personnel costs.
AI can be trained to generate polite, inclusive and personalized responses, maintaining a human tone even in automated communications.
This is just the first step. The possibilities for evolution of agentic HR systems are enormous and are already taking shape in cutting-edge companies:
Direct connection with Applicant Tracking Systems like Greenhouse, Workable or Lever, to automatically synchronize applications and status updates.
The system could access a corporate knowledge base to autonomously answer candidate questions about: benefits, selection processes, company culture, remote working policies.
Integration with Google Calendar or Calendly to automatically propose interview slots based on team availability and candidate preferences.
- Sergio Sentinelli, CTO AIONSoft
In our courses we show you how to build real AI automations that transform business processes. No theory, only practical projects.