AI for HR & Recruiting · Lesson 2
Resume Screening with Claude
Batch resume screening with a structured rubric in the prompt, extracting key criteria, scoring 0–10 with reasons, shortlisting workflow, and avoiding bias in criteria.
Practical exercise
What to do after this lesson
Take 10 real (or anonymized) resumes for an open position. Define three criteria blocks with weights. Use the lesson prompt to score each resume. Build a shortlist. Check: is there any systematic bias in the shortlist?
Ready-to-use prompt
Template for this lesson
Copy and adapt to your context. Text in angle brackets should be replaced.
You are an experienced HR recruiter. Evaluate this candidate's resume for the [JOB TITLE] role and provide a structured assessment.
SCORING CRITERIA:
- Technical skills (weight 40%): [LIST TECHNOLOGIES/SKILLS]
- 9–10: proficient in all key skills
- 7–8: proficient in 70%+, gaps are learnable
- below 6: critical gap in a key skill
- Relevant experience (weight 30%): minimum [N] years in [FUNCTION/INDUSTRY]
- Culture indicators (weight 30%): look for signals of [LIST: e.g., ownership, data-driven thinking, startup experience]
IMPORTANT: ignore name, age, photo, and university (unless it's in the criteria). Evaluate only on content.
Output strictly as JSON:
{
"score": number 0–10 (weighted average),
"breakdown": { "technical": number, "experience": number, "culture": number },
"strengths": ["specific strength 1", "..."],
"concerns": ["specific concern 1", "..."],
"recommendation": "shortlist" | "maybe" | "pass",
"one_liner": "one-sentence summary for the hiring manager"
}
CANDIDATE RESUME:
[PASTE RESUME HERE]