AI for HR Leaders: 6 Workflows to Automate This Quarter

HR is the single highest-leverage place to deploy AI inside a company. Every workflow — hiring, onboarding, performance, L&D, employee comms — is text-heavy, repetitive, and judgment-based. That's exactly what LLMs are good at. Here is the playbook I run with HR leaders in my corporate workshops.

The 6 workflows to automate first

1. JD writing — turn a 3-bullet brief into a polished, inclusive JD in 60 seconds.
2. Resume screening — score CVs against a rubric, not gut feel.
3. Interview question banks — role + competency + level → 10 calibrated questions.
4. Offer & rejection letters — on-brand, legally clean, in any tone.
5. Onboarding plans — 30/60/90 day plan personalised to the hire.
6. Performance review drafts — manager notes → STAR-format review.

Prompt: JD writer

Act as a senior talent partner. Write a job description for:
Role: {title}
Level: {IC3 / Manager / Director}
Must-haves: {3 bullets}
Nice-to-haves: {2 bullets}

Format:
- 2-line hook
- What you'll do (5 bullets)
- What we're looking for (5 bullets)
- Compensation philosophy line
Tone: confident, inclusive, no buzzwords.

Prompt: resume scorer

Score this resume against the JD on:
1) Relevant experience (0-5)
2) Skills match (0-5)
3) Trajectory (0-5)
4) Red flags
Return a markdown table + one-paragraph recommendation.
JD: """{jd}"""
Resume: """{resume}"""

Guardrails HR must set

• Never paste full PII into public chatbots — use enterprise tiers.
• Humans approve every outbound candidate communication.
• Document the rubric — the rubric is the policy, not the LLM.
• Audit shortlists for bias quarterly.

HR teams that adopt this get back 8–12 hours per recruiter per week — time that goes back into talent strategy, not paperwork.