Built an end-to-end recruitment screening system that processes candidate applications from submission through CV extraction, AI assessment, structured results, and CRM entry, reducing the manual work involved in reviewing and organizing large volumes of applications.
The challenge
WorkForceHub needed a more efficient way to process a high volume of candidate applications. With the potential to receive 500+ applications per month, manually reviewing CVs, comparing candidates against role requirements and entering applicant information into a recruitment tracker could create significant repetitive work and slow down the initial screening process. The objective was to automate the repetitive first stage of recruitment while keeping the final hiring decision with the recruitment team.
The solution
Designed and deployed an AI-powered candidate screening workflow that automates the process from application submission to structured candidate assessment and CRM entry. The system supports multiple recruitment roles and evaluates each applicant based on the specific position they applied for rather than applying a generic screening checklist.
Workflow diagram
System architecture
What I built
Candidates submit personal information, work experience, availability, role specific responses and CV through a structured form. Supports multiple recruitment roles with role specific screening for each position.
One intake point for every role.
The uploaded CV is retrieved automatically from the application submission. No manual downloading required.
CVs flow straight into the pipeline.
PDF.co extracts readable text from the candidate CV.
Makes CV content accessible to the AI screening engine regardless of file format.
The AI evaluates each application and CV against the specific requirements of the role the candidate applied for.
Each applicant is assessed individually based on their chosen position rather than a generic checklist.
The AI assessment is converted into consistent structured data.
Ensures every candidate record follows the same format regardless of AI output variation.
Each candidate receives a unique identifier automatically.
Enables consistent tracking and referencing across the recruitment process.
The complete candidate profile and screening results stored automatically including screening score from 0 to 100, recommendation of STRONG FIT, POTENTIAL FIT or NOT A FIT, key strengths, concerns and gaps, role specific assessment and supporting reasoning.
Recruiters get a structured shortlist-ready record for every applicant.
Implemented across CV retrieval, PDF extraction, AI processing, JSON parsing and CRM record creation. If a temporary failure occurs the affected operation retries automatically instead of stopping the workflow.
Designed to keep the automation reliable during continuous high-volume operation.
Automation engine
Operational impact
Measurable results
First-stage screening automated from submission to CRM record
Designed monthly application capacity
Screening consistency across every candidate and role
Screenshots
The full workflow from the Tally trigger through CV retrieval, PDF extraction, AI screening, JSON parsing, ID generation and Google Sheets record creation.
Purpose · Shows the complete screening engine in one view.
Candidate form capturing position, basic information and work experience.
Purpose · Demonstrates the structured multi-role intake experience.
Skills, qualifications, certifications and availability captured during the application.
Purpose · Shows the role-specific data collected for AI screening.
Google Sheets tracker with candidate IDs, contact details, positions and experience.
Purpose · Shows the structured records created for each applicant.
Google Sheets tracker with screening scores, recommendations, strengths, concerns and role-specific assessment.
Purpose · Shows the AI output recruiters review.
Make.com run history showing every step completed from application to CRM record.
Purpose · Confirms the workflow runs reliably end to end.
Key outcomes
A manual first-stage recruitment process where every CV had to be reviewed by hand, candidate data entered manually, and no consistent screening criteria existed across a growing volume of applications.
A Make.com scenario triggered by Tally submissions that retrieves CVs via HTTP, extracts text with PDF.co, screens each candidate against role specific requirements with Make AI Toolkit, structures results with JSON Parser, generates unique IDs with Make Tools and records everything in Google Sheets — with automatic error handlers and retry mechanisms across all critical stages.
The system evaluates each candidate against relevant experience, role specific skills, certifications and licences, availability, preferred work type, weekend availability, evening and night availability, application to CV consistency and overall suitability. Each candidate receives a screening score from 0 to 100 and one of three recommendations: STRONG FIT, POTENTIAL FIT or NOT A FIT, along with key strengths, concerns and gaps, a role specific assessment and the reasoning behind every recommendation. The workflow includes automatic error handlers and retry mechanisms across every critical stage — if a temporary service failure occurs during CV retrieval, PDF extraction, AI processing, JSON processing or CRM record creation the affected operation retries automatically rather than stopping the workflow. The architecture is designed to scale beyond 500 applications per month with practical capacity depending on the usage limits of the connected services. The finished system transforms candidate application, CV processing, AI screening, structured assessment and recruitment CRM entry into a single automated workflow — a scalable recruitment workflow built not just to automate the process but to remain reliable while running.
All data shown in screenshots is sample data used for demonstration purposes.