🤖 I automated candidate onboarding for a job assistance agency (from 45min to 30 seconds per file)

Nafiou Tino

2d • Wins 🌟

I recently completed an automation project for a client who runs a job search assistance service in Belgium.

❌ The problem before:

For each new candidate registered via a form, she had to manually:

Create a Google Drive folder with 3 subfolders
Download and organize the CV, cover letter, diplomas
Search for matching job offers on the Forem website (Belgian public employment service)
Compare the candidate's profile with available offers
Update a Google Sheets dashboard
Send a notification for follow-up

⏱️ Estimated time: 30 to 45 minutes per candidate

With 10-15 new registrations per week, that easily represented 6 to 10 hours of repetitive work every week.

✅ The solution:

A 33-node n8n workflow that does everything automatically as soon as a candidate submits the form:

Data reception → Webhook connected to Tally form
Smart matching → Double-filter algorithm that compares the profile to job offers (job title, location, skills, contract type...)
Client folder creation → Complete structure on Google Drive
Document organization → CV, letters, diplomas uploaded automatically
Dashboard updated → All info + links to matched offers
Notification → Telegram alert to inform that a new candidate is ready

🏆 Results:

Before: 30-45 min/candidate → After: ~30 seconds
Before: Manual job search → After: Automatic matching with scoring
Before: Risk of forgetting/errors → After: 100% standardized process
Before: 6-10h/week → After: Time freed up for human support

According to stats, automation can reduce candidate processing time by up to 50% and HR teams can save up to 15 hours per week.

🔧 Challenges encountered:

Tally data mapping → The form had 50+ fields with different formats (checkboxes, multi-select, files...). I had to create a "cleaning" node to normalize everything.
The matching algorithm → Finding the right balance between "too strict" (few results) and "too broad" (irrelevant results). I opted for a 2-level system: broad filter then precise scoring.
File management → Downloading files from Tally and uploading them to Google Drive with the correct naming in the correct folder.

💡 What I learned:

Automation isn't there to replace humans, but to free up their time for high-value tasks. My client can now focus on personalized support for her candidates instead of doing administrative work.

Questions about the project or n8n? I'll answer in the comments 👇

8 comments