Hiring

How it works

The mechanics of this kit, plainly — what runs and when, where the data lands, and what you look at.

• Hiring sweep runs mondays and thursdays · 8am and keeps the Hiring dashboard fresh. Mondays and Thursdays — moves each candidate forward from the evidence and names whose write-up is owed.
• What is captured lands in the Role register record, which lives at Data/Role Register in your vault — yours to open, search, and edit like any other file.
• What is captured lands in the Candidate pipeline record, which lives at Data/Candidate Pipeline in your vault — yours to open, search, and edit like any other file.
• What is captured lands in the Debrief log record, which lives at Data/Debrief Log in your vault — yours to open, search, and edit like any other file.
• The Hiring dashboard reads from the record and shows, at a glance: the pipeline, stage by stage · waiting on feedback · candidates on the record · what is coming · where they sit, by role · time in stage — median days · the offer on the table · the roles — tap one to scope the board · every candidate on the record · how this board works.

The wiring, in one breath: Birch holds Candidate capture · Birch holds Debrief capture · Birch holds Feedback nudge · Candidate capture writes to Role register · Candidate capture writes to Candidate pipeline · Hiring sweep writes to Role register · Hiring sweep writes to Candidate pipeline · Debrief capture writes to Candidate pipeline · Debrief capture writes to Debrief log · Role register shown by Hiring · Candidate pipeline shown by Hiring · Hiring kept fresh by Hiring sweep · Hiring sweep run by Birch.

Getting the most from it

A kit earns its place when it changes what you look at and what you say, not when it is installed. This is the rhythm that makes it pay.

• Read The pipeline, stage by stage after a sweep — Mondays and Thursdays — rather than mid-week: each band is how many people got at least that far, so the drop between two bands is where your process actually loses people. The ones who joined and the ones who did not sit outside the funnel, because a stage column cannot say which step someone left from.
• Treat Waiting on feedback as the panel that earns the kit. It ages from the day the person entered the stage and names the interviewer who owes the write-up — and when one has been sitting too long, ask Birch to chase it: Feedback nudge writes the reminder to your colleague, in your voice, into your drafts. It is never sent for you, and it is never addressed to the candidate.
• Say what happened the way you would say it to a colleague — "Alice is through to the panel on Thursday with Sam", "we made Bob an offer", "Nadia went elsewhere" — and Candidate capture moves the one row that changed. With two roles live, name which; if you do not, Birch asks rather than guessing.
• Get the debrief down while it is fresh: tell Birch what the interviewer said and Debrief capture writes it to the Debrief log in their words, with their own recommendation attributed to them, and marks that feedback as in. Birch never adds a verdict, never averages a panel, and never says who to hire — the record is there so the decision conversation has something better than memory.
• Use Time in stage — median days when a hire feels slow but nobody can say where: it is the median of the people sitting in each stage right now, so a tall bar is the step your process is actually stuck on. Tap a role in the rail to scope the funnel, the count, the bar and the calendar to that one hire.
• Keep the register honest at the two ends: open a role the moment you decide to hire (the sweep can only file people against roles it knows), and when a role fills, say so — it keeps its row and its history, so the funnel can still be read back next time you hire for it.

Make it yours

Nothing here is fixed. The pieces are ordinary agents, routines, records and dashboards, so anything you know about shaping those applies. Retire whatever earns nothing.

Things people usually change first

Say any of these in chat, or edit them on the kit’s page in the app.

• Which mailbox applications arrive at, so the sweep looks in the right place.
• The rounds each role runs and who sits on them — they are the role's own, in your words, so a designer's portfolio review and an engineer's technical live side by side.
• When the Hiring sweep runs, or pausing it between hires and keeping the pipeline by hand.
• How far back the first sweep looks, if three months is not the shape of this hire.
• An interviewer who should never be chased, and agencies whose mail should not be filed — standing preferences, not rows.
• The register is yours to edit: a stage corrected, a role closed, a candidate's row put back the way it was.

Your first run

Installing starts a short setup conversation — not a form. Answer roughly, correct later.

1. Installing ends at the first sweep's own screen, where the email and calendar access is asked for — that is the whole pitch, so the ask sits in front, where it makes sense. Start it and Birch reads back through three months of applicant and scheduling mail and the interviews already in your calendar, and rebuilds the pipeline: every candidate, the role they applied for, the stage the evidence puts them at, what is booked and who is on it. You watch the board fill. It is a real run and uses credits like any other.
2. When that first sweep finishes, Birch leaves one card with what it found — how many candidates, across how many roles — and an invitation to spend a few minutes fixing what it misread: which roles are actually open, anyone who is not a candidate at all, a stage it read wrong, and who interviews for each role. That is also where you say which mailbox applications arrive at, which lands in preferences rather than on a row.
3. Skip the sweep and the pipeline still works by hand: Birch takes the roles first — what you call each one, the team, roughly when it opened, the rounds it runs and who interviews — and then who is in play for each, one at a time. Nothing in that conversation touches your email, and the Hiring sweep waits on your Routines screen until you choose to run it.
4. After that the sweep runs on Mondays and Thursdays at eight, reads only what is new since the last one, moves a stage only where a message or a calendar entry says so, and leaves at most one short card: who is newly in, what is booked before the next sweep, and whose write-up has been owed longest. It reads your mail and your calendar; it never replies, never writes an event, and never contacts a candidate.

Watch the first pass, then tell it what you would have done differently — that correction teaches it more than any setup could.

The essentials

Learning, preferences and run costs work the same in every kit — so they are covered once, in the Kits guide.

Living with a Kit — the Kits guide →

← All kits