Fairy for People Ops.
Agents run the people workflows. Fairy judges what affects a person.
Onboarding, PTO, policy questions, and recruiting ops run on agents. The decisions that touch someone’s livelihood — comp, offers, terminations, performance, employee data — are held, judged, and routed to your People team, then a vetted external expert.
The problem
HR is high-volume — and every decision has a person on the other end.
Most people work is repetitive coordination that agents can absorb. But buried in that volume are the calls that carry real weight: a comp adjustment, an offer, a termination, a performance rating, a read of sensitive data. Get one wrong and it’s a discrimination claim, a wrongful-termination suit, or a data-privacy violation.
You can’t let agents run HR unsupervised. Fairy removes the bottleneck: agents run the volume, and every consequential decision is judged and routed to a person before it lands.
How Fairy governs it
- Compensation, offers, and equity — never decided by an agent alone
- Terminations, layoffs, and disciplinary action held for human sign-off
- Performance ratings and promotion decisions routed to managers + People
- Bulk employee-data reads gated for approval (PII / GDPR / SOC 2)
- Worker classification and immigration calls escalated to an expert
- Policy and benefits answers handled by agents, checked for accuracy
- Recruiting ops — sourcing, screening, scheduling — run at volume
- Onboarding / offboarding checklists executed with least-privilege access
- Every action logged to an audit trail with an accountable owner
- Employment-law questions cascaded to a vetted external expert
What’s at risk
People decisions carry legal and human weight.
Discrimination exposure
An agent screens candidates or sets comp with a biased pattern — creating EEOC exposure and reputational damage no one caught.
Wrongful termination
An automated action ends or alters someone's employment without the review, documentation, or legal check the situation required.
PII mishandling
An agent reads or moves bulk employee data — SSNs, salaries, health info — outside policy, triggering GDPR / privacy obligations.
Compliance violation
A worker-classification, immigration, or leave decision made without expertise fails an audit or an employment-law standard.
FAQ
Common questions.
What People Ops work can agents run through Fairy?
Agents handle the high-volume, low-stakes people work end to end: answering policy and benefits questions, processing PTO and expense requests, drafting job descriptions, screening and scheduling candidates, running onboarding checklists, and keeping HRIS records tidy. Fairy governs every one of these so nothing that affects a person's employment happens unsupervised.
Which people decisions does Fairy always route to a human?
Anything consequential to a person or the company's legal exposure: compensation changes, offer terms, terminations and layoffs, performance ratings, promotions, disciplinary action, immigration and classification calls, and any action that reads or moves sensitive employee data. Fairy's floor never lets an agent make these alone — they escalate to your People team, then to a vetted HR or employment-law expert.
How does Fairy protect employee data (PII) in HR workflows?
Every agent action runs through Fairy's governed gateway with scoped, least-privilege access to HR systems, and every action is logged to an audit trail. Actions that read bulk personal data or touch sensitive records are held for human approval by policy, giving you an accountable record for GDPR, EEOC, and SOC 2 obligations.
Can a handful of people run HR for a large company this way?
That's the model. Agents absorb the ticket volume and coordination; your People team and a network of external HR and employment-law experts are the backup for the decisions that carry legal or human weight. A small team sets the policy and catches the exceptions instead of doing the throughput.