Back to projects
Built with n8nHRautomation

Auto Recruiting Bot

From CV submitted to interview booked — zero human involvement

M

Built by Minahal Salahudin

Automation Engineer and Trainer

The Problem

Hiring teams read every CV manually, write individual emails, and track candidates in messy spreadsheets. Strong candidates wait days for a response. The team spends more time managing applications than evaluating talent.

How It Works

1

Form Polling

Google Forms polled every 30 minutes for new SWE and BDM applications

2

Normalization

Inconsistent field names from both forms mapped to a unified structure. Ghost rows and empty submissions filtered.

3

Duplicate Detection

Every candidate checked against master sheet and current batch — no duplicate processing

4

Resume Download

Drive URLs converted to direct download links and files fetched via OAuth

5

AI Screening

Groq (LLaMA) evaluates each CV with a role-specific prompt — returns score, classification, strengths, and concerns

6

Master Sheet Update

Full candidate record written to Google Sheets with AI results and status set to pending_email

7

Email Dispatch

Separate hourly workflow reads pending candidates and sends interview invites, hiring manager alerts, or rejection emails

8

Calendar Booking

Strong candidates automatically receive a Google Calendar interview invite

Key Features

Two Decoupled Workflows

Processing and emailing are split. If AI fails mid-run, no emails are sent for partially processed candidates.

Role-Specific AI Evaluation

SWE prompt evaluates technical depth and system design. BDM prompt evaluates revenue numbers and sales methodology.

3-Way Classification

Strong candidates get interview invites. Average candidates trigger a hiring manager review. Weak candidates receive a warm rejection.

Automatic Calendar Booking

Strong candidates receive a Google Calendar event with them added as attendee — no scheduling back-and-forth.

Duplicate Protection

Handles same-batch duplicates and resubmissions — no candidate is ever processed twice.

Results & Impact

  • Full screening pipeline runs automatically every 30 minutes
  • Interview invites sent and calendar booked without any human action
  • Hiring manager only sees pre-scored, pre-summarised strong candidates
  • Zero duplicate processing regardless of resubmissions
  • Works simultaneously for two completely different job roles

Time & Money Saved

Time Saved

5–8 hours per hiring cycle

Value Generated

Eliminates need for a dedicated HR coordinator for early screening

Tech Stack

n8n

Workflow engine

Groq + LLaMA 3

AI resume evaluation

Google Sheets

Application tracking and master database

Google Drive

Resume file storage

Gmail

Candidate and hiring manager emails

Google Calendar

Automatic interview scheduling

skillIT

Want this built for your business?

We build custom automation systems for companies across Pakistan and beyond. Tell us what's costing you time — we'll build the system that fixes it.

Get in Touch →