Skip to main content

AI HR assistant

Describe a role, paste a policy, or share your people data and get HR work you can act on: an interview plan and scorecard, an offer letter, a 30/60/90 onboarding plan, a performance review, a headcount plan. Zeplik works from the roles, policies, and data you give it, not a generic template, and it is honest about the line: it drafts the documents and the reasoning, and it tells you where employment law and your own counsel need to weigh in before anything is final.

Free to start. Nothing sends on its own.

Who it is for

How it works

  1. Share the role, policy, or people data

    Drop in the job description, the handbook section, the comp data, or the headcount numbers. Nothing is assumed. The work is built only from what you share, so the more real context you give, the more useful and less generic the output.

  2. Ask for the HR work you need

    Say it plainly, for example "design the interview for this role" or "draft an offer for Jane." Zeplik routes to the right people skill without you naming it, and you can steer the tone, the level, and the detail.

  3. Review, adjust, and get sign-off

    You get a ready-to-edit plan, letter, or review with the reasoning shown. Adjust it to your team, then run anything with legal or comp exposure past counsel and finance before it goes out.

What you can ask

Real asks in plain language, and the concrete deliverable you get back. Open the assistant to try your own; nothing runs until you send it.

Design the interview plan and questions for this role.

A competency-based interview plan for the role you describe: the loop, who covers what, question sets tied to each competency, and scorecards the panel can grade against, so the process is consistent and defensible.

Draft an offer letter for the candidate we picked.

An offer letter built from the details you paste, with the total comp package assembled (base, equity, signing bonus) and negotiation guidance for the hiring manager, ready for your review and counsel sign-off.

Build a 30-day onboarding plan for a new hire.

An onboarding plan with the pre-start checklist (accounts, equipment, buddy), a Day 1 and Week 1 schedule, and 30/60/90-day goals scoped to the role and team you describe.

Help me write a performance review for my report.

A manager review draft that turns your vague feedback into behavioral examples, with calibration prep and a structure you can edit, not a generic rating you have to defend from scratch.

Is this pay competitive? Model the comp band.

A comp analysis over the data you paste: band placement, outliers, and retention risks, with the equity grant modeled, so you can make an offer call that holds up.

Give me a headcount plan and a people snapshot.

An org and headcount plan with hiring sequence and team structure options, plus a leadership-ready people report on turnover, span of control, and flight risk from the data you provide.

What it can do

The assistant routes to the right skill for you. Each one below is a focused, ready-to-run workflow; follow a link to see exactly what it does.

Hiring and offers

Run the pipeline from open role to signed offer.

  • Hiring workflowChain the full hire: job post, pipeline, interview plan, offer, onboarding.
  • Interview designCompetency-based question sets and scorecards for the panel.
  • Offer lettersAssemble the total comp package and the offer letter text itself.
  • Pipeline trackingStage-by-stage funnel, stuck candidates, and conversion rates.

Onboarding and performance

Set new hires up and run review season.

Comp and org planning

Get pay and structure right before you commit.

People data and policy

Turn HR data and the handbook into clear answers.

  • People reportsTurnover, representation, span of control, and flight-risk readouts.
  • Policy answersPlain-language answers on PTO, benefits, travel, and leave.

What a general chatbot gets wrong on HR, and what is different here

A raw chatbot will write a confident offer letter or a policy answer that ignores employment law, your comp bands, and your actual handbook. Zeplik builds from the roles, policies, and data you give it, and it is honest that anything with legal or pay exposure needs your counsel and finance before it is final.

The failure. A general model drafts an offer or a termination note without a thought for at-will rules, protected classes, or state-by-state law.

How Zeplik grounds it. Zeplik drafts the document from your inputs and flags where employment law applies, so you know exactly what to run past counsel before anyone signs or receives it.

The failure. Asked about pay, a general model invents market numbers from memory and presents them as fact.

How Zeplik grounds it. It works from the comp data you paste, shows band placement and outliers against your own bands, and is clear it is not a live market feed, so you can validate before you make the offer.

The failure. A general model writes a review or a policy answer that sounds official but does not match your actual handbook or your review cycle.

How Zeplik grounds it. It builds from the policy and context you provide, keeps the reasoning visible, and stays a draft for your review, so the final wording is yours and matches your real process.

What it does not do

Knowing the boundary is part of using it responsibly. This is where a person, not the assistant, has to own the call.

This is HR drafting and analysis for your review, not legal advice or a compliant decision on its own. Employment, pay, and policy rules vary by jurisdiction and change often, so have qualified HR and legal counsel review anything that affects hiring, compensation, discipline, or termination before you act, and keep people data confidential.

Frequently asked questions

Is the AI HR assistant free?
Zeplik is free to start and every new account includes credits. HR skills run like any other Zeplik chat, so you spend only normal usage credits, with the cost shown before you send.
Can it write interview questions and scorecards?
Yes. Describe the role and it produces a competency-based interview plan with question sets tied to each competency and scorecards the panel can grade against, so the loop is consistent.
Can it draft an offer letter?
Yes. It assembles the total comp package from the details you paste and drafts the letter text with negotiation guidance for the manager. Have counsel review it before you send.
Is any of this legal advice?
No. It drafts HR documents and flags where employment law applies, but it is not a lawyer. Run offers, terminations, and policy language past your own counsel before they are final.
Does it connect to our HRIS or ATS?
Not yet. It works from what you paste and is honest about it. It does not pull Workday, Greenhouse, or Rippling, so bring the roles, pipeline, and people data you want it to use.
What do I need to give it?
For an interview, the role and level. For an offer, the candidate and package details. For onboarding, the role and start date. For a review, your notes on the person. For comp, paste the data.

More on Zeplik

Hire and manage with a faster first draft

Describe a role or paste a policy and get an interview plan, an offer letter, an onboarding plan, or a performance review you can act on. Free to start, and honest about what your counsel needs to check before it is final.

AI HR Assistant: Hiring, Onboarding, Performance & People Reports | Zeplik Chat