Short answer: The questions employees ask an HR chatbot cluster into eight predictable categories — leave and time off, payroll and expenses, benefits and insurance, onboarding, remote work and equipment, policies and compliance, training and career, and offboarding — and they add up to roughly 50 recurring questions at most companies. This guide lists all 50, tells you exactly what a correct answer needs to contain, and flags which questions a chatbot should never try to answer alone.
If you are building or refining an HR chatbot, the knowledge base is the part that actually determines whether employees trust it. A chatbot that can only handle a handful of generic questions gets abandoned within a week. One built around this list — with answer templates that state a fact, an effective date, and an escalation rule — holds up under real use.
This article is written for HR ops leads and People teams building the knowledge base themselves, not evaluating vendors.
TL;DR
- 50 recurring HR chatbot questions, grouped into 8 categories: leave, payroll, benefits, onboarding, remote work, policies, training, and offboarding.
- Each answer template needs one fact per paragraph, a stated effective date, and a clear escalation rule for edge cases.
- Some questions should never be answered by the chatbot alone — disciplinary matters, harassment reports, medical details, and salary disputes always route to a human.
- GDPR compliance for an HR chatbot comes down to three practical things: EU hosting, no personal employee data in the knowledge base, and role-based access to conversation logs.
- Loading this list into a RAG-based HR chatbot takes five steps and, with Heeya, under 30 minutes once your documents are ready.
Table of Contents
- Leave & Time Off Questions
- Payroll & Expense Questions
- Benefits & Insurance Questions
- Onboarding & First Weeks Questions
- Remote Work & Equipment Questions
- Policies & Compliance Questions
- Training & Career Questions
- Offboarding Questions
- How to Write Answers a Chatbot Can Use
- Questions an HR Chatbot Must Never Answer Alone
- GDPR and Data Protection for an HR Chatbot
- Loading These Questions Into a RAG Chatbot in 5 Steps
- Measuring Whether It's Working
- Further Reading
- FAQ
Leave & Time Off Questions
Leave questions are the highest-volume category for most HR teams, usually asked right before someone wants to book something. Each answer needs the current accrual rule and a pointer to where the request gets submitted.
| Question | What the Answer Must Contain | Source Document |
|---|---|---|
| How many PTO days do I have left this year? | Accrual method, where the live balance is displayed (HRIS), not a fixed number | Leave policy |
| How do I request time off and who approves it? | Submission tool, required notice period, approver hierarchy | Leave policy |
| Can I roll over unused vacation days into next year? | Carryover cap, deadline to use rolled-over days, forfeiture rule | Leave policy |
| What is our sick leave policy — do I need a doctor's note? | Days allowed without documentation, threshold that triggers a note requirement | Leave policy / handbook |
| How much notice do I need to give for planned leave? | Standard notice window, exceptions for emergencies, who to notify first | Leave policy |
| What happens to my PTO balance if I resign? | Payout rule (varies by location), timing relative to final paycheck | Leave policy / offboarding checklist |
| Do we offer parental or family leave, and how do I apply? | Eligibility window, duration, pay structure, application lead time — flag as escalation-adjacent | Leave policy / benefits guide |
Payroll & Expense Questions
Payroll questions spike around pay dates and reimbursement deadlines. Employees want dates and process — keep these answers short.
| Question | What the Answer Must Contain | Source Document |
|---|---|---|
| When does payroll run and when does direct deposit hit? | Pay frequency, exact processing dates, typical bank posting delay | Payroll schedule |
| Why does my paycheck amount look different this period? | Common causes (holiday pay, tax bracket change, benefits deduction) plus escalation to payroll for account-specific review | Pay stub explainer |
| How do I update my direct deposit or tax withholding info? | Exact self-service tool, form name, processing lead time | HRIS user guide |
| How do I submit an expense reimbursement request? | Submission tool, required receipt format, approval chain, reimbursement timeline | Expense reimbursement policy |
| What is the maximum I can expense for travel or meals? | Per-category caps, pre-approval threshold, receipt requirement | Expense reimbursement policy |
| Why was a deduction taken from my paycheck? | List of standard deductions (benefits, retirement, garnishments) plus escalation for anything unrecognized | Pay stub explainer |
Benefits & Insurance Questions
Benefits literacy is consistently low, not because the information is missing but because it lives in a dense plan document nobody reads end to end.
| Question | What the Answer Must Contain | Source Document |
|---|---|---|
| What health insurance plans are available and what do they cost? | Plan tiers, employee cost share, where to compare side by side | Benefits guide |
| When is open enrollment and can I change plans outside of it? | Open enrollment window (exact dates), qualifying life events that allow off-cycle changes | Benefits guide |
| Does my plan cover dental and vision, and what's the deductible? | Coverage inclusion by plan tier, deductible amount, in-network provider search link | Plan summary |
| How do I add a dependent to my health coverage? | Enrollment window for dependents, required documentation, submission tool | Benefits guide |
| What is the difference between our HSA and FSA options? | Contribution limits, rollover rules, employer contribution if any | Benefits guide |
| Does the company match retirement contributions, and how much? | Match formula, vesting schedule, enrollment deadline | Benefits guide |
Onboarding & First Weeks Questions
New hires generate the highest question density of any employee lifecycle stage — and are least likely to interrupt a colleague to ask something they suspect they "should already know." An always-available chatbot removes that hesitation.
| Question | What the Answer Must Contain | Source Document |
|---|---|---|
| What do I need to bring or complete on my first day? | ID/documents checklist, forms to complete before arrival, first-day schedule | Onboarding checklist |
| Where do I find the org chart and who is my manager's manager? | Link to the current org chart tool, note on how often it's updated | Onboarding checklist |
| How do I set up my company email, VPN, and required software? | Step-by-step setup guide, IT contact for issues | IT setup guide |
| What is the dress code, and does it change for client meetings? | Standard dress code, client-facing exceptions | Employee handbook |
| When do my benefits become active after I start? | Waiting period length, effective date rule, what's covered during the gap | Benefits guide |
| Who do I contact if my equipment hasn't arrived yet? | IT/facilities contact, expected turnaround, loaner equipment option if any | Onboarding checklist / IT setup guide |
| What training is mandatory in my first 30, 60, and 90 days? | Required course list by milestone, deadline for completion, where to access the LMS | Onboarding checklist / LMS catalog |
Remote Work & Equipment Questions
Hybrid and remote arrangements generate a steady stream of logistics questions, mostly procedural rather than policy debates — among the easiest to fully automate.
| Question | What the Answer Must Contain | Source Document |
|---|---|---|
| Am I eligible to work remotely, and how many days per week? | Eligibility criteria by role, standard remote allowance, exceptions process | Remote work policy |
| What equipment does the company provide for a home office? | Standard equipment list, stipend amount if applicable, request process | Remote work policy |
| How do I request a laptop repair or replacement? | Ticketing tool, expected turnaround, loaner policy | IT setup guide |
| What is the policy on working from another country temporarily? | Maximum duration allowed, approval requirement, tax/compliance flag for escalation | Remote work policy |
| Do I need to be online during specific core hours? | Core hours window if any, time zone expectations for distributed teams | Remote work policy |
| Who do I contact for IT support outside office hours? | After-hours contact channel, response-time expectation, urgent vs. non-urgent triage | IT setup guide |
Policies & Compliance Questions
Handbook questions are where a chatbot outperforms a static document most visibly: instead of scanning 80 pages, an employee gets the exact section in one exchange. Keep this category strictly informational — point to policy, never interpret it for a specific situation.
| Question | What the Answer Must Contain | Source Document |
|---|---|---|
| Where can I find the full employee handbook? | Direct link/location, note on last revision date | Employee handbook |
| What is the policy on workplace conflicts of interest? | Definition, disclosure requirement, who to notify | Code of conduct |
| What is the process for reporting a safety concern? | Reporting channel, anonymity option if available, expected response timeframe | Employee handbook |
| Can I hold a second job while employed here? | Outside-employment policy, disclosure requirement, conflict-of-interest boundary | Employee handbook |
| What is our social media and confidentiality policy? | What can and can't be shared publicly, confidential information categories | Code of conduct |
| How does the company handle data privacy for employee information? | What data is collected, where it's stored, who has access, contact for privacy requests | Privacy notice |
| What are the consequences of violating the code of conduct? | General framework only — flag as escalation-adjacent, never a specific determination | Code of conduct |
Training & Career Questions
Career questions are lower volume but high stakes for retention. A chatbot won't replace a career conversation with a manager, but it removes the friction of finding the process.
| Question | What the Answer Must Contain | Source Document |
|---|---|---|
| What courses are available through our learning platform? | Catalog overview, how to enroll, whether courses are self-paced | LMS catalog |
| Does the company reimburse tuition or certification costs? | Annual cap, eligible programs, pre-approval requirement, reimbursement timing | Tuition reimbursement policy |
| How often are performance reviews conducted? | Review cycle frequency, timing, where results are documented | Employee handbook |
| How do I request a transfer to a different team or role? | Internal mobility process, minimum tenure requirement, who to notify first | Internal mobility policy |
| What is the process for requesting a raise or promotion? | General process and timing window only — always route the actual conversation to the manager | Employee handbook |
| Are there mentorship or internal mobility programs available? | Program overview, how to apply or express interest | Internal mobility policy |
Offboarding Questions
Departing employees ask fewer but more anxious questions during an emotionally charged transition. Keep these answers factual and calm — not a place for a chatbot to improvise.
| Question | What the Answer Must Contain | Source Document |
|---|---|---|
| When will I receive my final paycheck after I leave? | Final pay timing rule, what it includes (accrued PTO, outstanding expenses) | Offboarding checklist |
| How does continued health coverage work after I resign? | Continuation option overview, enrollment deadline, cost structure — flag for benefits team follow-up | Offboarding checklist / benefits guide |
| What happens to my unused PTO when I leave? | Payout rule, how it's calculated, when it's paid | Offboarding checklist |
| Who do I return company equipment to, and by when? | Return process, deadline, consequence of non-return if applicable | Offboarding checklist |
| What happens to my unvested equity or retirement account after I leave? | General forfeiture/vesting rule only — always route specifics to a human | Offboarding checklist / plan documents |
How to Write Answers a Chatbot Can Use
The 50 questions above are only half the work. What actually determines whether an employee trusts the answer — and comes back the next time — is how the answer is written. Three rules do most of the heavy lifting.
One fact per paragraph. A source document written for legal review often buries the actual answer inside three paragraphs of context. A chatbot answer should do the opposite: lead with the fact, then add context only if it changes the answer.
State the effective date. HR policies change — plan years reset, accrual rates get revised, remote work rules get updated. An answer with no date attached looks authoritative right up until it's wrong. Every template should carry a line like "Effective since [date]," updated whenever the underlying policy changes.
Build in an escalation rule. Not every version of a question has a documented answer. "How many vacation days do I get?" is answerable from policy. "Can I take vacation during our busiest month if my manager already said no?" is not — it needs judgment your documentation doesn't cover. Write each template with an explicit fallback: if the retrieved passage doesn't fully cover the question, the agent should say so and point to a named contact rather than guess.
A weak template says: "You can request time off through the portal." A strong one says: "Request time off in [HRIS name] under Time Off > Request, with at least [X] business days' notice. Current as of [date]. For anything outside this, contact [HR contact]." Same fact, but the second version survives contact with a real employee.
Questions an HR Chatbot Must Never Answer Alone
Some employee questions should never reach a knowledge base at all — not because the technology can't retrieve an answer, but because the situation requires human judgment, discretion, or documentation that a chatbot cannot provide safely.
Disciplinary and performance issues. Any question touching an active disciplinary process or termination decision needs a human with full context and the authority to speak to it. A chatbot guessing at process here creates real legal exposure.
Harassment and discrimination reports. These require immediate, confidential human handling — never a bot interaction. Route these instantly and visibly to a named contact or reporting channel.
Medical and health details beyond general benefits information. A chatbot can explain what a health plan covers in general terms, but it should never interpret a specific medical situation or handle protected health information.
Salary negotiation and compensation disputes. Pay is individual and often confidential. The chatbot can describe the general review process; it should not discuss specific numbers.
Immigration and visa status questions. These carry legal consequences that vary by circumstance — always route to whoever manages immigration cases at your company.
In practice, this means configuring your agent's system guidance with explicit refusal rules: if a message contains signals like "disciplinary," "harassment," "medical," "salary," or "visa," the agent should stop and hand off to a named human contact rather than attempt an answer.
GDPR and Data Protection for an HR Chatbot
An HR chatbot processes employee data, which makes it subject to data protection law wherever your employees are based. In the EU, that means the General Data Protection Regulation (GDPR). The practical requirements are narrower than they sound — three things cover most of what matters for a document-based HR chatbot.
EU hosting. If your employees are in the EU, conversation data and the knowledge base should stay on EU infrastructure, or be transferred under a valid legal mechanism. A vendor that can't say clearly where the data lives is not ready for an HR deployment. Heeya stores all agent data on EU infrastructure by default. See our guide on data sovereignty for AI chatbots in the EU.
No personal data in the knowledge base. This is the most overlooked control. Feed the chatbot policy documents — the leave policy, the benefits guide, the handbook — not individual employee records or exported HRIS data. All 50 questions above are answerable from policy documents alone.
Role-based access to conversation logs. Employee conversations may reveal sensitive context even when the chatbot doesn't store personal data long-term. Limit who can view raw conversation history to the HR roles that need it, and set a retention period.
Transparency. Employees should know they're talking to an AI, not a person, from the first message — good practice, and a requirement under EU AI transparency rules. Document this in your handbook.
France's CNIL publishes practical guidance on AI systems that process personal data, a useful reference beyond the regulation text. For general HR process guidance, SHRM is a widely used resource. None of this replaces your own legal counsel's review — this section is a starting checklist, not legal advice.
Loading These Questions Into a RAG Chatbot in 5 Steps
Once you have answer templates for your version of these 50 questions, getting them into a working HR chatbot is mostly a document exercise, not a technical one. Here's the sequence that works with a RAG-based platform like Heeya.
- Step 1 — Create the agent and write its scope. Name the agent, then write a short system guidance statement: what it should answer, what it should never touch, and where to send anything outside its scope.
- Step 2 — Turn your answer templates into source documents. Organize the 50 questions by category — leave, payroll, benefits, and so on — as sections in your handbook or standalone plain-language FAQ documents. One topic per section improves retrieval accuracy.
- Step 3 — Upload the documents. Heeya accepts PDF, Word, PowerPoint, Excel, and plain text files, plus website crawling. Each file is chunked and indexed automatically.
- Step 4 — Configure escalation and the contact form. Enable the contact form tool so the agent captures an employee's details for human hand-off, and confirm the hard-stop triggers from the "never answer alone" list are in the system guidance.
- Step 5 — Deploy the widget and test with real questions. Embed the chat widget with a single line of code, then run the agent through real employee questions — including a few that should trigger escalation — before rolling it out.
For a 200-person company with an already-organized document set, this typically takes well under 30 minutes of setup time. The variable is document prep, not the platform.
Measuring Whether It's Working
Once the chatbot is live, two numbers tell you almost everything you need to know: how much it's actually handling, and what it's failing to handle.
Containment rate. This is the share of employee questions the chatbot resolves without a human follow-up. Track it by category, since policy lookups contain far more easily than anything touching individual circumstances. A low containment rate in a category means either the documentation is thin or the questions genuinely need a human.
Unanswered-question log. A RAG platform worth using shows you the questions the agent couldn't answer confidently. Review this weekly at first, then monthly. Each unanswered question is either a knowledge-base gap to fix or a signal the question belongs on the "never answer alone" list.
Heeya's built-in conversation analytics surface both of these directly, so you're not exporting chat logs into a spreadsheet to figure out what to fix next.
Ready to build your HR chatbot's knowledge base?
Upload your policy documents, set your escalation rules, and deploy an HR assistant trained on these 50 questions — flat monthly pricing, EU-hosted, no credit card required to start.
Start free — no credit card required View pricingFurther Reading
- AI Chatbot for HR: Employee Support Automation in 2026 — HR chatbot use cases, deflection benchmarks, and HRIS integration patterns.
- HR Chatbot vs Intranet FAQ: Why Employees Skip It — why static FAQ pages underperform a conversational interface.
- AI Chatbot for Employee Questions — a deployment-focused guide for HR Directors and People Ops leads.
- Employee Onboarding AI Agent — a playbook for the first-90-days question surge covered above.
- AI Chatbot Data Sovereignty in the EU — the full explainer on EU hosting requirements.
FAQ
What questions can an HR chatbot answer?
A well-configured HR chatbot can answer the roughly 50 recurring questions that fall into eight categories: leave and time off, payroll and expenses, benefits and insurance, onboarding, remote work and equipment, policies and compliance, training and career, and offboarding. These are all questions with a documented, factual answer. Questions requiring judgment — disciplinary matters, harassment reports, medical details, salary negotiation, or immigration status — should always route to a human HR contact instead.
Is an HR chatbot GDPR compliant?
An HR chatbot can be GDPR compliant when it follows three practical controls: EU data hosting (or a valid transfer mechanism) for employee conversation data, a knowledge base built from policy documents rather than personal employee records, and role-based access to conversation logs with a defined retention period. Employees also need to be told clearly that they are interacting with an AI. Heeya stores agent data on EU infrastructure by default and provides a Data Processing Agreement on paid plans.
How many questions should an HR chatbot knowledge base cover?
Most companies find that 40 to 60 distinct questions, grouped into 6 to 8 categories, cover the large majority of routine employee contact volume. The 50-question list in this guide — spanning leave, payroll, benefits, onboarding, remote work, policies, training, and offboarding — is a practical starting point. Coverage matters less than accuracy: a smaller set of well-sourced, dated answers with clear escalation rules performs better than a large set of vague ones.
What is the difference between an HR chatbot and a standard FAQ page?
A standard FAQ page requires the employee to find and read the right entry themselves, using whatever wording the page author chose. An HR chatbot built on Retrieval-Augmented Generation (RAG) understands natural-language questions phrased however the employee asks them, retrieves the relevant passage from your actual documents, and answers directly — without the employee having to scan a static page.
How do I turn these questions into a chatbot knowledge base?
Write an answer template for each question using three rules: one fact per paragraph, a stated effective date, and a clear escalation path for anything the documentation doesn't fully cover. Organize the templates by category, either as sections in your existing handbook or as standalone documents, then upload them to a RAG-based chatbot platform. Each document is automatically chunked and indexed for retrieval.
What happens when an employee asks a question outside the chatbot's scope?
A well-configured agent should recognize when a question falls outside its documented scope or touches a sensitive category — disciplinary, medical, harassment, salary, or immigration topics — and hand off to a human contact instead of guessing. This is typically configured through the agent's system guidance and a contact form tool that captures the employee's details for HR follow-up.
How often should HR chatbot answer templates be updated?
Review the knowledge base on a quarterly cycle at minimum, and immediately whenever an underlying policy changes — a new plan year, a revised leave policy, an updated payroll schedule. Each answer template should carry an effective date, so outdated content is easy to spot during review rather than surfacing as a wrong answer to an employee first. — Written by Anas Rabhi.
Stop answering the same HR questions by hand.
Deploy an HR chatbot trained on your own policies — flat monthly pricing, EU-hosted, live in under 30 minutes. No credit card required to start.
