AI for HR & People

Why AI matters in HR

A 1-person HR function with AI runs like a 4-person team. Sourcing, screening, onboarding doc generation, performance review drafts, internal comms — all handled with prompts and tools. The benchmark: 3-5x recruiter productivity, 30-50% faster time-to-hire, onboarding plans in minutes not days.

The trap: AI in HR is a regulated minefield. Bias in screening, discrimination in scoring, EEO compliance. Build governance from day one.

The core HR AI stack

Recruiting

• Gem — recruiting CRM with AI sourcing.
• Paradox — Olivia conversational recruiter.
• hireEZ — sourcing across 800M+ profiles.
• SeekOut — diversity and deep-skill sourcing.

People ops + payroll

• Rippling — HR, IT, payroll on one platform.
• Gusto — best SMB payroll + benefits.

Performance

• Lattice AI — performance reviews with AI assist.
• 15Five — weekly check-ins, manager coaching.

Doc generation

• Claude for job descriptions, offer letters, review drafts.

Deep dive 1: AI-assisted recruiting

The workflow:

1. AI-drafted JD. Claude or Jasper writes the first draft from a role brief. Manager edits.
2. Sourcing. Gem or hireEZ searches LinkedIn + public profiles for fit.
3. AI screening. Paradox or a custom prompt asks initial qualifying questions.
4. Interview prep. AI generates question bank from the JD + candidate resume.
5. Decision support. AI synthesizes interviewer feedback into a hiring recommendation.

The non-negotiable: AI cannot make the hire/no-hire call. Humans decide; AI informs.

Quick start prompts: Job description writer, Candidate screening rubric.

Deep dive 2: Onboarding at SMB speed

Most SMB onboarding is a Notion doc and a prayer. AI does better.

The workflow:

1. 30-60-90 plan generated from role description and team context.
2. Stakeholder map — who they meet week 1-4.
3. Manager checklists auto-generated.
4. First-week schedule built in calendar.

Quick start prompt: 30-60-90 onboarding plan.

Deep dive 3: Performance reviews

Manager dread peak. AI cuts review-writing time by 80% while improving quality.

The workflow:

1. Pull check-in notes, peer comments, project outcomes for the employee.
2. AI drafts the review covering strengths, development areas, goals.
3. Manager reviews, adds personal touches, finalizes.

The line: AI drafts. Manager writes the final. Never auto-send a performance review.

Quick start prompt: Performance review draft.

Governance

• EEO and bias. AI scoring of candidates is regulated (NYC AEDT, Illinois, EU AI Act). Audit your tools for bias. Document the criteria.
• Disability accommodation. AI screening must offer reasonable accommodation paths.
• Data privacy. Candidate data is regulated (GDPR, CCPA). Vendor data controls non-negotiable.
• Performance review. Never auto-finalize. Always human-reviewed.
• Layoff communications. AI can draft. Humans deliver. Never auto-send difficult news.

30-60-90 day plan

Days 1-30: deploy AI for JD writing and screening rubrics. Pilot on one role. Measure time-to-fill.

Days 31-60: add sourcing automation (Gem or hireEZ). Build the prompt library — 8-10 saved prompts.

Days 61-90: add performance review drafting. Onboarding automation. Internal comms templates.

Maturity model

• Level 1: HR person uses ChatGPT for drafts.
• Level 2: AI in sourcing, screening rubrics, review drafts. Bias audit done.
• Level 3: Conversational AI for candidate screening at scale. AI-assisted onboarding.
• Level 4: Predictive people analytics. AI identifies flight risk, manager effectiveness, team health.

Recommended prompts

• Job description writer
• Candidate screening rubric
• 30-60-90 onboarding plan
• Performance review draft
• Internal change comms

Where to go next

• Browse HR & People tools →
• For an AI assistant that drafts your internal comms and handles scheduling, see the AI Chief of Staff playbook.