High review volume
Recruiters need a faster way to organize the first review without turning the product into a full applicant-tracking system.
CodelyHub Product Case Study
ShortlistLens: AI CV screening platform for recruiters and candidate ranking.
This AI CV screening platform for recruiters was built by CodelyHub to help hiring teams complete the first CV review more systematically by connecting role setup, candidate CV uploads, background screening, evidence-based results, credit handling, and completion notifications in one workflow.
The Challenge
Why an AI CV screening platform for recruiters needs structured evidence.
The product was designed around a practical recruitment problem: one role can attract dozens of CVs, and manually opening every file makes it difficult to review candidates consistently and explain why one person appears stronger than another.
Evidence matters
A score alone is not enough. The screening flow needs to show the skills, experience, gaps, and evidence that support the result.
Background processing
Uploaded CVs need to move through screening, storage, credits, status updates, and notifications without forcing the user to manage each technical step.
The Product Flow
A clear four-step journey from role setup to candidate review.
The customer-facing workflow keeps the experience simple while background services handle the screening process.
01
Create the role
A hiring project represents one position and stores the role requirements used for screening.
02
Upload candidate CVs
Candidate files are uploaded to the project and queued for background processing instead of being handled one by one manually.
03
Run background screening
Supabase triggers the production queue worker, n8n claims the work, OpenAI supports CV analysis, and results are saved back to the database.
04
Review results
The dashboard presents completed screening results for review and export, while customer and owner notifications confirm job completion.
Production Architecture
The screening engine is event-driven rather than a manual button-to-script demo.
A production database webhook connects the website and Supabase screening queue to the n8n worker. Once a screening job is created, the system can claim queued work, process CVs, save results, update job status, handle credits, and send completion emails.
Supabase
Authentication, projects, screening jobs, stored results, credits, and database-triggered events form the application data layer.
n8n Queue Worker
The production workflow claims queued screening items, runs processing logic, saves completion data, retries failures where appropriate, and handles notifications.
OpenAI + deterministic logic
AI assists with extracting and evaluating candidate evidence, while the product is positioned as decision support and leaves the final hiring decision to people.
Verified Production Test
The full event-driven screening path completed successfully.
A production test confirmed the complete sequence from website job creation to Supabase insert, database webhook, n8n queue claim, OpenAI screening, database completion, and both customer and owner emails. A verified one-CV run completed with one file processed and zero failed files.
1 CV completed
The screening result was saved successfully and the job status was confirmed as completed.
0 failed
The verified test finished without a failed CV item.
End-to-end notifications
The completed result appeared in the customer dashboard after refresh, and both customer and owner completion emails were received.
Responsible Screening
The product supports the first review; it does not replace the hiring decision.
ShortlistLens is designed to surface candidate evidence, matched requirements, gaps, and ranking information so a recruiter can review the strongest candidates first. Human review remains part of the decision process.
Evidence-based output
Results are intended to show matched skills and experience, missing requirements or weak evidence, and the rationale behind a screening result.
Project-level organization
Each hiring role remains its own project so candidate screening and rankings stay connected to the correct job requirements.
Credit-aware processing
The backend tracks reserved, used, and refunded credits so failed or retried processing can be handled without silently losing usage balance.
Technology
A SaaS product connected across website, database, automation, AI, and email.
ShortlistLens combines a custom web application with Supabase, n8n, OpenAI, private CV storage, background queue processing, and completion notifications.
Supabase + Web App
User authentication, hiring projects, CV uploads, screening jobs, stored results, and account credits.
n8n + Webhooks
Event-driven queue processing, retries, database updates, cleanup, orchestration, and completion notifications.
OpenAI
AI-assisted screening works within the product workflow to help analyze candidate information against the role requirements.
What This Project Demonstrates
Product engineering and automation working as one system.
This project demonstrates more than a standalone n8n workflow. It connects a customer-facing SaaS interface with authentication, storage, database events, background processing, AI analysis, usage controls, notifications, failure handling, and saved screening results.
Need a workflow-backed product?
Connect your website, database, AI, and business operations into one system.
CodelyHub AI can help design and automate products that need background processing, AI-assisted analysis, database workflows, notifications, and reliable API integration.