AI for recruiters
Recruiting is one of the easiest places to misuse AI. Generated cold emails, generated JDs, generated screening questions all start to read the same — exactly what kills response rates and candidate experience. The recruiters winning with AI are using it for context-loading and pattern-recognition, then bringing their own judgment to the parts candidates notice.
What AI handles well
Job descriptions a candidate would actually read
The problem. Most JDs sound like compliance docs. "Rockstar," "fast-paced environment," "must wear many hats." Strong candidates self-select OUT.
What AI does. Use the JD prompt. It writes a JD around what the person will actually DO, the honest tradeoffs, what success looks like, and a real comp range — not generic boilerplate.
Sourcing outreach that doesn't look like a template
The problem. Passive candidates get 12 LinkedIn DMs a day. They're all "Hi {first name}, I'd love to connect." Yours blends in. You get a 5% reply rate.
What AI does. Use the LinkedIn DM prompt. Reference the specific thing about THEM, make a small ask, no template-smell. Higher reply rate from fewer messages.
Cold outreach to senior candidates
The problem. You're sourcing for a senior IC or director role. They have options. Generic "great opportunity!" emails go straight to trash.
What AI does. Use the cold outreach prompt with their LinkedIn + your role context. Get 3 angles to test. Lead with a specific observation about them, not a generic compliment.
Screening question generation
The problem. You're screening for a role you've never hired for. Generic "tell me about a challenge" questions don't reveal anything.
What AI does. Use AI to generate role-specific screen questions: behavioral, technical, motivational. Output should test the actual things the role requires, not generic resilience.
Candidate debrief structure
The problem. You did a panel. Now you have to debrief 4 interviewers, each with their own opinions. The debrief drifts to whoever talks loudest.
What AI does. Use the meeting summary prompt to structure the debrief: rubric scores, behavioral evidence, dissent, decision. AI keeps the conversation grounded in what was observed, not vibes.
Rejection emails that aren't soulless
The problem. You have to reject 12 candidates this week. The default template is the worst part of recruiting. The candidate deserves better.
What AI does. Use AI to draft rejection emails that reference something specific from their interview, give honest (not vague) feedback, and leave the door open for future fit. Not every reject needs detail — but the strong-but-not-this-role ones do.
Reference check questions that surface real signal
The problem. Reference checks are theater. "Were they a good employee?" "Yes." "Would you hire them again?" "Yes." You learn nothing.
What AI does. Use AI to generate behavioral reference questions tied to your concerns from the interview process. "Tell me about a time when X" instead of "are they good?"
Compensation research synthesis
The problem. You have offers from 3 sources, internal benchmarks, and a candidate's expectation. They don't agree. You need to make a recommendation.
What AI does. Have AI structure the comp data (range, percentile, equity assumptions) and surface where the disagreement is. Output: a defensible recommendation with the reasoning behind it.
Your AI stack
Start with the foundation. Add specialized tools as the work calls for them.
Foundation LLM
Specialized add-ons
Prompts ready to use
Get started in 30 minutes
Run the JD prompt on the next role you're hiring for
15 minCompare to your old JD. Notice what's gone (rockstar, fast-paced) and what's added (real work, honest tradeoffs, comp). Ship the new version.
Build a Claude Project per role you commonly hire for
10 minLoad: the JD, the rubric, your top-5 must-haves, common red flags. Every prompt about candidates for this role is now pre-loaded with context.
Set up AI meeting notes for your next 3 interviews
5 minStop typing during calls. Listen. The AI captures what was said. You evaluate after.
Common mistakes
AI-generated outreach that all sounds the same. Defeats the entire point of personalization. Candidates can spot template-smell instantly.
AI-screening candidates without understanding what it's optimizing for. AI scores against patterns it learned — which can quietly perpetuate bias. Audit what it's actually selecting for.
Using AI to bypass the relational part of recruiting. AI helps you PREPARE for the conversation. It can't replace the trust and rapport that close offers.
Pasting confidential candidate info (resumes, salary expectations, internal feedback) into public AI. Use enterprise/team tiers, anonymize, or skip.
Generic AI-rejection emails to every candidate. The strong-but-not-this-role candidates deserve specifics — they're also future hires, future referrers, future advocates.