Muse for Recruiters: Sourcing Research, Candidate Briefs, and a Pipeline That Doesn't Go Quiet
A recruiter's working setup for Muse: research a role before you source it, brief every candidate before the call, and sweep the pipeline weekly so no one…
By Professor Muse8 min read
Recruiting is three recurring jobs that never finish. Research the role before you source it, brief yourself before every candidate call, and keep the pipeline moving so nobody waits in silence. None of them is hard on its own. Together they eat the week, because each one starts with the same chore: open the job description, the hiring manager’s notes, the resumes, and the message thread, and remember where each candidate actually stands.
That is the work Muse is good at in a recruiting seat. It reads the material you give it, pulls the relevant parts together, and turns them into drafts. It leaves the judgment calls to you: who advances, what you offer, what goes out under your name. Meta’s rule for Muse is that nothing publishes, sends, or spends without your approval, and that rule is what makes this setup safe to run on a pipeline full of real people’s livelihoods. Muse drafts, you decide and send.
The short version
- Connect calendar, inbox, and messages first, read-only, and keep approval required for every outreach message, interview invite, and rejection until the drafts have earned your trust.
- Run three jobs: sourcing research before you open a role, a candidate brief before every call, and a Friday pipeline sweep of everyone waiting on you.
- Every claim in a brief needs a source: the resume line, the public profile, or the hiring manager’s note it came from. A claim without a source is a guess wearing a suit.
- Nothing sends without your approval. That is Meta’s rule for Muse, and it is what keeps a delegated inbox from becoming a delegated voice to candidates.
The setup: The Recruiting Desk
Call the whole arrangement the Recruiting Desk. I run it for a recruiter carrying four open roles at once, where the week runs on a fixed rhythm. Monday is sourcing research for the role that opens next. Wednesday is candidate briefs for the week’s calls. Friday is the pipeline sweep. Each job is a scheduled routine, and each one lands as a document to read, not as a message sent on your behalf.
The Recruiting Desk has three jobs, and each one maps to a prompt below.
Sourcing research. Before you message anyone, Muse reads the job description and the hiring manager’s notes and turns them into a sourcing brief: the three must-have skills, the two nice-to-haves that are not must-haves, and the places people with that profile actually show their work in public. A sourcing brief written from the job post alone will chase the posting’s wish list. A brief written from the hiring manager’s notes chases what the team will actually hire for.
The candidate brief. Before every call, Muse reads the candidate’s resume and public profile and writes a one-page brief: what they have done that maps to this role, the two questions worth asking first, and anything in their background that needs a careful question rather than an assumption. The brief is preparation only. You make the hiring judgment on the call.
The pipeline sweep. Every Friday, Muse reads the pipeline and lists everyone waiting on you: who is owed a reply, who finished an interview and has heard nothing since, and which hiring manager has sat on feedback past the day you promised it. For each one, Muse drafts the next message. You read every draft before it goes out, because a follow-up sent to the wrong candidate is worse than a follow-up sent late.
| Job | Muse does | You do |
|---|---|---|
| Sourcing research | Turns the job description and hiring manager notes into a sourcing brief | Decides the must-haves, approves where to look, writes the first outreach |
| Candidate brief | Reads the resume and public profile, drafts the brief and opening questions | Runs the call, judges the answers, decides who advances |
| Pipeline sweep | Lists everyone waiting, drafts the next message for each | Reads every draft, sends what is true, chases the hiring manager yourself |
Before any of this runs, do the permissions pass from our Sentinel permissions guide: approval required for every send, and read-only access everywhere else until the drafts have earned your trust. The same draft-then-human-send rule our sales lesson uses for follow-ups applies here to every candidate message. And the privacy rules apply to sourcing: public profiles and material candidates sent you are the inputs, and a candidate’s private contact details, salary history, or anything you would not say to their face stays out of prompts.
Our sales lesson is the closest cousin to this desk, because prospecting and sourcing run on the same pattern: research first, brief before the conversation, follow up before the trail goes cold:
Prompt 1: the sourcing brief
Run this when a role opens, before you message anyone about it.
Write a sourcing brief for [role title] from [job description / hiring manager notes]. Give me: the three must-have skills, the two nice-to-haves that are not must-haves, the one thing that would rule a candidate out, and three places people with this profile show their work in public. Quote the hiring manager’s note for each must-have. Do not write outreach yet. Brief only. I approve the brief before any sourcing starts.
Without the must-have versus nice-to-have split, a sourcing brief will treat every line of the job post as equal, and you will spend the week screening for a skill the team would have traded away on day one.
Prompt 2: the candidate brief
Run this before every candidate call, once you have the resume and the public profile.
Write a one-page candidate brief for [candidate name] for [role title] from [resume / public profile]. Cover: what they have done that maps to this role, with the resume line or profile link for each point; the two questions worth asking first; and one thing in their background that needs a careful question rather than an assumption. Flag anything you could not verify instead of filling it in. Draft only. I run the call and make the judgment.
A brief that fills gaps with plausible detail is the dangerous kind, because it reads like preparation and behaves like a rumor. A brief that flags what it could not verify tells you exactly where to ask.
Prompt 3: the Friday pipeline sweep
Run this every Friday afternoon, against the pipeline as it stands.
Sweep my pipeline for [the four open roles]. List everyone waiting on me: who is owed a reply, who finished an interview and has heard nothing since, and which hiring manager feedback is past due. For each candidate, draft the next message in under 80 words, addressed by name, stating exactly where they stand. Do not send anything. Draft only. I read and send every message myself.
A sweep that only lists candidates and skips the hiring managers misses half the bottleneck. Feedback past due is usually the reason a candidate is waiting, and naming it on Friday is what gets it moving on Monday.
What can go wrong: the inflated brief
Here is the failure mode you will meet first. The candidate brief comes back polished: the candidate “led the redesign” of a product, “owned” a hiring pipeline, “drove” a launch. It reads like a strong match for the role. Then you open the resume, and the truth is narrower. The candidate worked on the redesign team, contributed to the pipeline project, and supported the launch. Muse upgraded verbs, and the upgrade changed the story from contributor to owner.
If that brief had gone into the call unchallenged, you would have opened with questions that assumed a seniority the candidate never claimed, and the candidate would have spent the call correcting your premise instead of showing their actual work. Worse, if the brief had been forwarded to a hiring manager, the inflation would have travelled under your name. Approval fatigue lets it through: on a Wednesday packed with calls, a brief that reads well gets trusted on the strength of its prose, and the one upgraded verb rides along with it.
The fix has two parts. First, Prompt 2 already requires the resume line or profile link for each mapped point, and you hold it to that: a point without a source gets sent back. Second, before every call, check the one load-bearing claim yourself: open the resume and confirm the candidate’s actual role in the work you plan to ask about. It takes a minute. That check is the recruiter’s version of the rule in our failure modes lesson: Sentinel gates what Muse does, like sending. Verifying a brief is your job, every time.
Start with one job
Do not build the whole Recruiting Desk this week. Pick the role that opens next, run Prompt 1, and read the brief against the hiring manager’s notes: did it split must-haves from nice-to-haves, did it quote the note for each must-have, did it invent a requirement nobody stated? If the brief survives that reading, add candidate briefs for next week’s calls, then the Friday sweep.
Muse does what it is good at in recruiting when you run it that way: the gathering, the first draft, the watching. You keep what was always yours: the judgment, the conversation, the send button. That gives you a pipeline where every candidate gets a brief before the call, every waiting candidate gets an answer on Friday, and nobody’s story gets upgraded on its way to a hiring manager. The edge is showing up prepared, honest, and on time, every week.
