AI CV and Resume Screening Workflow

Use AI to screen CVs and resumes against role requirements, identify top candidates based on a structured criteria matrix, and generate shortlist summaries for hiring managers. Reduce screening time while improving candidate quality consistency.

Transformation

Before & After AI

What this workflow looks like before and after transformation

Before

HR teams spend 6-8 hours per role manually reviewing 100-200 CVs, with inconsistent evaluation criteria across reviewers. Strong candidates are missed due to fatigue during high-volume screening. Unconscious bias influences shortlisting decisions. Hiring managers receive unstructured candidate summaries that make comparison difficult.

After

AI screens 200 CVs in under 30 minutes against a structured criteria matrix, surfacing the top 15-20 candidates with consistent scoring. Every candidate is evaluated against the same criteria regardless of when their CV arrives in the inbox. Hiring managers receive standardized comparison summaries with clear rationale for each shortlisted candidate, reducing time-to-shortlist by 75%.

Implementation

Step-by-Step Guide

Follow these steps to implement this AI workflow

1

Define Role Requirements and Must-Have Criteria

1-2 days

Work with the hiring manager to translate the job description into a structured screening criteria matrix. Separate absolute must-haves (dealbreakers) from nice-to-haves (differentiators). Assign weights to each criterion based on role priorities.

Role Requirements to Screening Matrix Converter
You are an HR screening specialist. Convert the following job description for [ROLE TITLE] at [COMPANY] into a structured screening matrix. Separate must-haves from nice-to-haves. Assign weights (1-10) to each criterion. Include: technical skills, experience level, education, certifications, soft skills, and cultural fit indicators. Output as a scoring rubric.
Have the hiring manager review the matrix before screening begins. Agreement on criteria upfront prevents disagreements at the shortlist stage.
2

Create a Screening Criteria Matrix

2-3 days

Build a detailed scoring matrix that maps each criterion to specific CV evidence markers. Define what "meets," "partially meets," and "does not meet" looks like for each requirement. Include bias-reduction guidelines to ensure fair evaluation.

CV Evidence Marker and Scoring Guide
You are a talent assessment expert. For the [ROLE TITLE] screening matrix, create a detailed evidence guide. For each criterion, list: (1) what to look for on the CV (keywords, achievements, certifications), (2) red flags to watch for, (3) how to score equivalent experience from non-traditional backgrounds. Include a section on cultural and regional CV format differences for SE Asian candidates.
Share this guide with everyone on the screening team to ensure consistent evaluation standards across reviewers.
3

Build AI Screening Prompts for Batch Processing

3-5 days

Create structured prompts that evaluate each CV against your criteria matrix. Design the prompt to output consistent, comparable scores with rationale for each criterion. Include instructions for handling edge cases like career changers, non-traditional backgrounds, and multilingual CVs.

CV Screening and Scoring Prompt
You are an HR screening assistant. Evaluate the following CV for the [ROLE TITLE] position against these criteria: [LIST CRITERIA WITH WEIGHTS]. For each criterion, score 1-5 with a one-sentence rationale. Calculate the weighted total. Flag any must-have gaps. Output: overall score, strengths, concerns, and a recommend/maybe/reject verdict.
Process one CV at a time for best accuracy. Save outputs to a spreadsheet for comparison across candidates.
4

Batch-Process Applications and Track Results

3-5 days

Set up a systematic workflow for processing all applications through the AI screening prompt. Track scores in a comparison spreadsheet, handle edge cases through manual review, and maintain an audit trail for compliance and fairness documentation.

Batch Screening Workflow and Tracker Template
You are an HR operations specialist. Design a batch CV screening workflow for processing [NUMBER] applications for [ROLE TITLE]. Include: (1) a spreadsheet tracker template with columns for each criterion score, (2) a process flowchart from application received to shortlist delivered, (3) quality control checkpoints, (4) compliance documentation requirements. Account for PDPA data handling in SE Asian markets.
Process in batches of 20-30 CVs per session to maintain quality. Take breaks between batches to avoid AI prompt fatigue.
5

Generate Shortlist Summaries for Hiring Managers

1-2 days

Compile the screening results into a structured shortlist report for the hiring manager. Include candidate comparison tables, individual summaries with strengths and concerns, and recommended interview focus areas based on screening gaps.

Shortlist Summary Report Generator
You are an HR business partner. Create a shortlist report for the hiring manager reviewing [ROLE TITLE] candidates. Include: (1) executive summary of the applicant pool, (2) comparison table of top [NUMBER] candidates scored against criteria, (3) individual candidate profiles (150 words each) with strengths, concerns, and recommended interview questions. Rank candidates by weighted score.
Send the report 24 hours before the shortlist review meeting so the hiring manager can prepare questions in advance.

Get the detailed version - 2x more context, variable explanations, and follow-up prompts

Tools Required

AI writing assistant for CV analysis (e.g., ChatGPT, Claude, Gemini)Spreadsheet tool for scoring tracker (e.g., Google Sheets, Excel)Applicant tracking system or shared document platformPDF reader or text extraction tool for processing CV formats

Expected Outcomes

Reduce CV screening time from 6-8 hours to under 1 hour per role for 200 applications

Improve screening consistency with structured criteria, reducing evaluator bias

Deliver standardized shortlist reports that enable faster hiring manager decisions

Cut time-to-shortlist by 75%, keeping top candidates engaged before competitors

Solutions

Related Pertama Partners Solutions

Services that can help you implement this workflow

Common Questions

AI-assisted screening is generally permissible, but you must comply with local data protection laws such as Singapore PDPA, Malaysia PDPA, and Thailand PDPA. Key requirements include obtaining consent for data processing, not making fully automated decisions without human oversight, and maintaining records of how decisions were made. Always have a human reviewer validate AI recommendations before sending rejection or shortlist notifications.

Three safeguards help. First, use structured criteria defined before screening begins, not after seeing the candidates. Second, anonymize CVs during screening by removing names, photos, and institution names where possible. Third, audit your results periodically by checking whether the AI scores correlate with protected characteristics. If you notice patterns, adjust your criteria or prompts to focus more tightly on job-relevant skills and achievements.

Transparency is both ethical and increasingly required by regulation. Best practice is to include a brief note in your job posting or application confirmation that AI tools assist in the initial screening process. Emphasize that all shortlisting decisions are reviewed by a human. This builds trust and positions your company as thoughtful about technology adoption.

AI handles multilingual CVs well if you specify the expected languages in your prompt. For SE Asian candidate pools, expect CVs in English, Bahasa, Thai, Vietnamese, and occasionally Mandarin. Use AI to translate and standardize before scoring. For non-standard formats (portfolios, video CVs, LinkedIn profiles), create separate evaluation prompts tailored to that format and map the output to your standard scoring matrix.

Ready to Implement This Workflow?

Our team can help you go from guide to production — with hands-on implementation support.