Overview
What This Solves
Hiring teams waste hours manually reviewing CVs, writing candidate notes, and comparing profiles. This playbook documents the AI agent pattern Alhasan built — automating candidate discovery, CV analysis, and structured summary generation using Claude.
How It Works
Step 1: Candidate Collection
The agent searches for candidates across defined sources (job platforms, internal referrals, LinkedIn exports) based on the job description and role criteria you define.
Step 2: CV Ingestion & Analysis
Each candidate CV is passed to Claude with a structured analysis prompt:
You are a senior talent acquisition specialist.
Analyse this candidate profile for the following role:
Role: {role_title}
Key Requirements: {requirements}
Candidate CV:
{cv_text}
Provide a structured summary with:
1. Role Fit Score (1-10) with brief rationale
2. Key Strengths (3-5 bullet points)
3. Experience Gaps or Risks (2-3 points)
4. Recommended Interview Focus Areas (2-3 questions)
5. One-line shortlist recommendation: RECOMMEND / HOLD / DECLINE
Keep each section concise. Focus on what matters for this specific role.
Step 3: Comparative Review
The agent aggregates all candidate summaries into a comparison table, ranked by fit score. Hiring managers review structured evidence — not raw CVs.
Results
- CV review time reduced from 3 hours to 25 minutes per batch
- Shortlisting decisions made on structured evidence, not first impressions
- 94% of interview invites went to candidates the agent rated 7+/10
Tools Required
- Claude API (claude-opus-4-8 or claude-sonnet-4-6 recommended)
- A CV parsing tool (any PDF-to-text converter)
- Optional: n8n or Make.com for workflow automation
Privacy Note
Ensure candidate CV data is handled in compliance with applicable data protection regulations. Use anonymised identifiers in the agent workflow where possible.