Recruiting

Claude for Recruiters: Your Candidate Pipeline, Searchable in Plain English

Recruiting is memory-intensive. You're tracking 40 candidates across 12 open roles, each at a different stage, each with their own constraints. Claude can help you think through candidates, draft outreach, and prep for interviews — but only if it knows what you know. And Claude forgets everything the moment the conversation ends.

Stash gives Claude a persistent, searchable record of your pipeline. Candidate profiles, role specs, interview notes, stage history — stored once, retrieved instantly.

What recruiters store in Stash

Candidate profiles: current role, skills, salary expectations, availability, referral source
Open roles: must-haves, deal-breakers, hiring manager preferences, headcount
Interview notes: what came up, red flags, strengths, next steps
Pipeline status: stage, date, outcome, who's waiting on what
Hiring manager briefs: their communication style, what they actually care about vs. the JD

Start of day: pipeline review

context() + search("final-round")

Claude loads your standing context (who you are, which desk you run) then returns everyone currently in final rounds. You see: three candidates across two clients, one with an offer pending since Tuesday. You know exactly where to focus before your first call.

Candidate intake

After a first call with a candidate:

add(collection="candidates", record={
  "name": "Aisha Patel",
  "current-role": "Senior Data Engineer, FinTech",
  "skills": "Python, Spark, dbt, Snowflake",
  "salary-expectation": "£90-100k",
  "availability": "1 month notice",
  "motivation": "wants a product company, currently in services",
  "notes": "Strong communicator, asked good questions about data culture",
  "status": "screening"
})

Later: search("Python dbt available") — Claude returns every candidate matching that profile. You're building a searchable talent library, not a spreadsheet you forget to update.

Matching candidates to new roles

A new role comes in: senior data engineer, fintech background preferred, must have dbt experience.

search("data engineer dbt fintech")

Claude surfaces Aisha plus any other candidates you've logged. You're not starting from memory — you're searching a record store. The recruiter who can surface the right person in 30 seconds wins the placement.

Interview debrief notes

add(collection="interviews", record={
  "candidate": "Aisha Patel",
  "role": "Senior Data Engineer - TechCorp",
  "date": "2026-06-07",
  "interviewer": "Sarah (Hiring Manager)",
  "outcome": "strong yes — loved the dbt deep-dive",
  "concern": "wants async culture, TechCorp is meeting-heavy",
  "next": "verbal offer pending comp discussion"
})

Now every debrief is searchable. When TechCorp asks you in a week "remind me what we said about Aisha" — you have it.

Staying on top of hiring manager preferences

Hiring managers are harder to predict than candidates. Log what you learn:

add(collection="hiring-managers", record={
  "name": "James Okafor — VP Engineering, TechCorp",
  "preference": "prefers candidates from scale-up backgrounds, not enterprise",
  "pet-peeve": "candidates who haven't looked at the product",
  "decision-speed": "fast if strong, ghosts if uncertain — follow up at 72h"
})

Limitation to know: Stash is keyword search, not AI ranking. It returns records that match the words you use. It won't score candidate fit or rank results — that's Claude's job once the records are retrieved.

Setup: two minutes

Sign up at stashlite.com with your Google account — free
Copy your connector URL, paste into Claude → Settings → Connectors
Log your first candidate or role in conversation

Free tier: 2,500 records, 50 searches per month. An active recruiting desk fits comfortably.

Build the memory your recruiting practice needs

Stash is free to start. Your pipeline is searchable in 2 minutes.

Get your connector URL →

Stash is a new service — pricing may change, cancel anytime.