An AI HR & recruiting agent that screens, schedules, and onboards.

What can an AI HR agent do?

A Unistaff AI HR agent screens inbound applicants against your criteria, schedules interviews on your calendar, chases documents, and runs onboarding paperwork — 24/7. It works to defined KPIs, keeps a full history, and asks a human to approve before anything is sent.

• Screening & shortlisting — reads applications and CVs, scores them against your must-haves, and shortlists the best fits.
• Interview scheduling — proposes slots, books on Google Calendar, sends invites, and handles reschedules.
• Candidate comms — answers applicant FAQs, sends status updates, and keeps warm candidates from going cold.
• Onboarding paperwork — collects documents, chases the missing ones, and kicks off day-one checklists.
• HR intake & FYIs — routine "how many PTO days do I have / where's the policy" questions, deflected from your desk.

No unread applicants. Every CV read the same day, against the same rubric, while you sleep.

How does an AI recruiting agent screen candidates?

It ingests your role criteria and scorecard, reads each application, ranks candidates against your must-haves, and shortlists the strongest — flagging edge cases for a human. It applies the same rubric every time, so screening stays consistent instead of drifting with fatigue.

1. Brief the role. You give it the job criteria, a scorecard and any deal-breakers — in plain language, the same way you'd brief a recruiter.
2. Read every application. It parses each CV and application against your must-haves, grounded in your instructions (RAG grounding), not a generic web model.
3. Score and shortlist. It ranks candidates against the scorecard and surfaces the top matches, with the reasoning attached so you can check its work.
4. Flag the edge cases. Anything ambiguous — a non-obvious background, a borderline score — goes to a human instead of a silent auto-reject.

This is the three-step setup behind every Unistaff role — a base language model, the recruiting behaviour and scorecard on top, then your data and criteria layered in. It reads for what you actually hire on, not keyword bingo.

Can it schedule interviews automatically?

Yes. It reads your calendar availability, proposes and books interview slots, sends the invites, and handles reschedules on Google Calendar and email — no back-and-forth. You set the rules for when and how long; it fills the slots.

• Reads availability — checks your Google Calendar for open, eligible slots.
• Proposes and books — offers the candidate options, confirms, and puts it on the calendar with an invite.
• Handles reschedules — cancellations and time changes get rebooked without you touching your inbox.

You approve the rules. It fills the calendar.

What does it do for onboarding?

It collects and chases onboarding documents, runs a day-one checklist, answers new-hire questions from your handbook, and logs everything. The paperwork grind that eats a hiring manager's first week gets handled in the background.

• Document collection — requests the forms and IDs you need, and chases the ones that don't come back.
• Day-one checklist — triggers the standard steps so nothing gets missed before a new hire starts.
• New-hire FAQs — answers "where's the policy / who do I ask about payroll" from your handbook, grounded in your docs.
• Full history — every request, reminder and answer is kept in memory and auditable.

Because it holds context in memory, it remembers who still owes a signed form and follows up until it lands. Nothing sits in a half-finished folder.

What does it connect to?

It plugs into the tools you already run: Google Calendar and Gmail for scheduling and candidate comms, Google Sheets or HubSpot / Pipedrive for the candidate pipeline, and Slack or Telegram to keep your team in the loop. Setup is no-code.

• Scheduling & comms — Google Calendar, Gmail
• Candidate pipeline — Google Sheets, HubSpot, Pipedrive
• Team loop — Slack, Telegram
• Payouts (contractor roles) — Stripe, where relevant

A candidate the agent shortlists lands in your pipeline. An interview it books shows up on your calendar. The work goes where you already look, not into a separate silo you have to check.

How is a human kept in control?

Human-in-the-loop by default. The agent drafts and proposes — a shortlist, an invite, a rejection — and you approve or override before anything is sent. Every action is recorded in an auditable history, and hiring decisions stay with people.

• Approve before it sends — no offer, rejection or message goes out without a human sign-off where you want one.
• Auditable history — every screen, score and message is logged and reviewable.
• Fair, consistent screening — the same scorecard is applied to every candidate, and edge cases route to a person rather than a silent auto-reject.

We built digital employees, not black-box chatbots, precisely because hiring is high-stakes. You approve. It sends.

How fast can it go live, and what does it cost?

Most roles go live in a day, with no code — you pick the HR role, brief it, and connect your calendar and pipeline. Pricing is per digital employee per month, from $149/mo, priced against a routine hire's salary rather than a seat.

• Live in a day — select the role, brief it in plain language, connect Google Calendar and your pipeline.
• Priced against a salary — a fraction of a loaded recruiting hire for the same routine workload, with no per-seat or per-token surprise billing.
• Paid 14-day pilot — run the HR role on one open position, credited to your first invoice.

See the tiers and the per-employee pricing and ROI vs hiring. Never off the clock, and a fraction of the cost.

AI HR agent vs. hiring a recruiter

An AI HR agent handles the routine screening, scheduling and onboarding grind 24/7 and to a consistent rubric; a human recruiter brings judgment, negotiation and the final hiring call. In practice they pair — the agent clears the queue, the human decides.

The cost row is a "fraction of a salary" frame, not an audited percentage — a full-time recruiting hire runs thousands a month, fully loaded. The judgment row is deliberately even: the final call is a human's.

What can't it do yet?

It doesn't make the final hiring decision, handle sensitive or legally weighted HR calls, or run delicate negotiation on its own. Those stay with people. The agent clears the routine screening, scheduling and paperwork so your judgment goes where it counts.

• Final hire / reject calls — it recommends and shortlists; a human decides who gets hired.
• Sensitive & legal HR — terminations, grievances and compliance-heavy decisions stay with a person.
• Negotiation nuance — offer and salary negotiation is drafted for a human, not run autonomously.
• Judgment on culture fit — it surfaces evidence; the read on a person is still yours.

We'd rather name the boundary than oversell past it. The routine is the agent's; the judgment is yours.

Frequently asked questions