Every AI use case, mapped to the business.

32 use cases owners are being sold this year, sorted by where they act in a business of 5 to 100 people. Only 3 are proven in production.

Status
Pick a room to see its use cases
How we rate

Four statuses, one rule: rate the evidence, not the pitch.

Each entry gets a status for how proven it is in businesses of 5 to 100 people, and a separate rating for how strong the evidence behind that status is.

Evidence strength

StrongSeveral independent studies or large independent surveys.
ModerateAt least one independent study or credible survey, plus consistent case evidence.
ThinA handful of independent data points.
Vendor claimsOnly what sellers publish about their own products.

How an entry is made

Each entry is drafted from public sources: academic studies, regulators, independent surveys and reputable reporting. Every figure carries a numbered citation you can open. Anything a vendor says about its own product is labelled as a vendor claim.

Review and updates

Every source in this edition was opened and checked against the text in October 2026. Each entry is then signed off by a named operator and re-checked when the evidence moves. Until that sign-off, the entry is marked as a draft at the top of the page.

The full guide

All use cases at a glance

Field GuidePeople & payrollAI onboarding assistants

AI onboarding assistants

Also called new hire chatbot, onboarding bot, AI buddy for new starters, induction assistant

EvidenceThin
ClassGenerative

What the evidence says about AI onboarding assistants for new staff, what they need to work, what they cost, and the paperwork they cannot replace.

ReviewDraft, awaiting operator review
Drafted27 October 2026
Fact-checkedEvery source opened, October 2026
Key takeaway

Onboarding is a real problem: only 12% of US employees say their company does it well, and new hires take around 12 months to reach peak performance⁠. The best independent evidence that AI helps comes from a call centre, where agents with two months' tenure and an AI assistant matched colleagues with more than six months without one⁠. We found no independent study of AI onboarding assistants themselves. Where know-how lives in people's heads rather than documents, the assistant has little to work from.

At a glance

StatusUnproven
What it doesGuides new starters through their first weeks: answers how-do-we-do-this-here questions, walks through checklists and drafts role-specific training plans from your own documents.
Typical costUS$21 per user per month, billed annually, for an office-suite AI add-on, cut to US$18 for the first year under a promotion running to 31 December 2026⁠. A team chatbot plan is US$25 per user per month, billed annually⁠ (vendor list prices, checked Oct 2026). Dedicated onboarding platforms are mostly quote-based.
Data it needsWritten procedures, role descriptions and an onboarding checklist; for paperwork, the new hire's personal, tax and banking details.
Setup effortMedium: the software is easy, but most small firms must first write down how the work is actually done.
EvidenceThin. Analogous studies of AI assistants helping newer workers in large firms; no independent study of onboarding assistants themselves.

What is an AI onboarding assistant?

It is a chatbot, or a feature inside HR software, that supports a new employee in their first weeks. It answers practical questions (where is the leave form, who approves purchase orders, how do we price a rush job), reminds them of checklist tasks and turns your documents into a short training plan for their role. Some versions also handle paperwork: collecting ID, tax and bank details, sending documents for signature and requesting system accounts.

The case for it is simple. New starters are reluctant to ask the same question twice, and owners are too busy to answer it. Gallup reports that only 29% of new hires feel fully prepared and supported to excel after onboarding⁠. An assistant that answers at any hour is an appealing patch.

How does it work?

Most onboarding assistants work like other document chatbots. The tool searches your procedures, handbook and role notes for relevant passages, and a language model writes an answer from them. Office-suite assistants do this over files the new employee is allowed to open⁠.

Better setups add structure: a checklist per role with due dates, a named human buddy and set check-ins at the end of the first week and month.

The assistant cannot invent your know-how. If the right way to cost a job, load a truck or handle a difficult client exists only in a senior employee's head, the assistant will either say it does not know or produce a plausible generic answer. Out-of-date and duplicated documents cause the same problem: in one large company's HR chatbot, near-duplicate versions were the main reason search found the wrong article⁠.

What it needs to work

An onboarding assistant can only repeat what you have written down. It needs written procedures for the role, a checklist with owners and dates, and a secure route for the personal information new hires hand over.

  1. A written procedure for each core task in the role, reviewed within the past year and kept in one shared folder.
  2. An onboarding checklist per role: accounts, equipment, policies to sign, training, and who owns each step by when.
  3. A single current employee handbook and current policies, with old versions archived where the assistant cannot read them.
  4. Contract templates and written particulars that meet local law: section 29 of the Basic Conditions of Employment Act in South Africa⁠, the day-one written statement in the UK⁠.
  5. A secure home for ID, tax and bank details, normally the payroll system, not chat history or email threads.
  6. A named manager or buddy for each new starter, with set check-ins at the end of the first week and the first month.

Most use cases in this guide succeed or fail here, on the data underneath them, before any AI is involved.

What does the evidence show?

We found no independent study of AI onboarding assistants as a product category. The claims for them come from vendors. The independent evidence is indirect.

The strongest comes from a study of more than 5,000 customer-support agents at one company⁠. An AI assistant helped newer and less skilled agents most. Agents with two months' tenure and AI access performed as well as agents with more than six months without it, and staff turnover fell as newer workers stayed⁠. That is one call centre, not a small firm.

A field experiment with 758 knowledge workers, run with Boston Consulting Group, found AI users completed 12.2% more tasks and worked 25.1% faster on tasks within the AI's abilities⁠. On a task outside them, they were 19 percentage points less likely to reach a correct answer⁠. New starters are the people least able to tell which kind of task they are facing.

AI onboarding assistant vs manager-led induction

Compared onAI onboarding assistantManager-led induction
AvailabilityAny time, for repeat questionsLimited to the manager's free time
Quality of answersOnly as good as your written proceduresDraws on experience and context
Relationship and cultureCannot replace a welcomeManager involvement is linked to onboarding success⁠
Cost modelPer-user subscription⁠Manager and colleague time
Best forDocumented processes and frequent hiringSmall teams and undocumented know-how

Status and evidence

StatusUnproven
Evidence strengthThin

Rated unproven because we found no independent study of AI onboarding assistants; the case for the category rests on vendor claims. The nearest independent evidence shows an AI assistant sped up newer agents in one call centre⁠, while AI misleads users on tasks beyond its abilities⁠, so the evidence is thin.

Where does it go wrong?

  • Your documents are thin or stale, so the assistant gives generic advice instead of how your business actually works⁠.
  • A new starter trusts a confident answer on a task the AI handles badly, and does the job wrong without knowing⁠.
  • The assistant replaces the human welcome. Gallup found employees were 3.4 times as likely to call their onboarding a success when their manager took an active role⁠.
  • Legal paperwork slips. In the UK the principal written statement of employment particulars is due on the first day of employment⁠. In South Africa written particulars are due when the employee starts and, if the employee cannot understand them, must be explained in a language and manner they understand⁠. A chatbot nudging the new hire does not discharge the employer's duty.
  • Personal details collected through a chat end up in the wrong place, or the new hire is never told what is collected and why⁠.

Numbers it should move

Staff cost ratioPayroll as a share of sales.
Response timeHow long a customer waits for a first reply.

Ask the demo business

Northgate Supply is a fictional wholesale distributor of catering and packaging supplies: 28 staff, R46.6m in sales over the last year. Its data is generated for this demo, to 30 September 2026.

Fictional data
What share of our sales goes on staff costs?
12.6%of sales over 12 months

Payroll took 12.6% of sales over the year. December's payroll was nearly double a normal month because of 13th cheques, but record sales held that month's ratio to 14.9%, about the same as February 2026, the quietest month for sales, at 14.7%. March's increases lifted monthly payroll by about 5.5%.

OctNovDec14.9%JanFebMarAprMayJunJulAugSep
How this was worked out

Monthly payroll cost, including 13th cheques, divided by monthly sales.

Answered in 2 msScanned 4 546 rowsFrom payroll, invoices
Want this on your own numbers, in the AI tools you already use?Book a demo

Where the data usually lives

Google WorkspaceMicrosoft 365SimplePayPaySpaceEmail inboxesExcel and Google Sheets

Where the records this use case needs are usually kept in a small business. It is not a list of recommended tools.

Frequently asked questions

How much does an AI onboarding assistant cost?

You can start with tools you may already pay for. An office-suite AI add-on lists at US$21 per user per month, billed annually, cut to US$18 for the first year under a promotion running to 31 December 2026⁠. A team chatbot plan lists at US$25 per user per month, billed annually⁠. Both were checked in October 2026. Dedicated onboarding platforms usually quote on request. The bigger cost is the time it takes to write the procedures the assistant needs.

Do I need a developer to set up an onboarding chatbot?

No. A shared folder of procedures plus an office-suite or team chatbot is enough to start. Before you do, check who can open which files, because these assistants answer from anything the user can already access⁠. Linking the assistant to payroll or e-signature systems is where outside help usually comes in.

Can AI help new employees get up to speed faster?

In one large study, yes: call-centre agents with two months' tenure and an AI assistant performed as well as colleagues with more than six months without one⁠. But on a task outside the AI's abilities, workers using it were 19 percentage points less likely to get the right answer⁠. It helps most where work is repeatable and already written down.

What do POPIA and UK GDPR mean for collecting new hire details with AI?

Onboarding collects ID numbers, tax and bank details and sometimes health information. POPIA requires you to tell people what you collect, why and who will receive it, and it treats health and biometric information as special personal information with stricter rules⁠. In the UK, the ICO's employment records guidance covers what to tell workers, retention and security⁠. Keep these details in your payroll system, not chat logs. This is not legal advice.

Can a chatbot handle the employment contract and paperwork?

It can remind and explain, but the duty stays with you. In the UK the employer must give the principal written statement on the first day of employment, with the wider statement within two months⁠. In South Africa the Basic Conditions of Employment Act requires written particulars when work starts, kept for three years after employment ends⁠. If the employee cannot understand them as written, they must be explained in a language and manner the employee understands⁠.

More in people & payroll

Sources

  1. 1
    Generative AI at Work
    arXiv (Brynjolfsson, Li and Raymond), 2024Study
  2. 2
  3. 3
  4. 4
  5. 5
  6. 6
    Basic Conditions of Employment Act 75 of 1997
    South African Government, 1997Government
  7. 7
    Protection of Personal Information Act 4 of 2013
    South African Government, 2013Government
  8. 8
    Employment practices and data protection: keeping employment records
    Information Commissioner's Office (UK), 2025Government
  9. 9
  10. 10
  11. 11
    ChatGPT Pricing
    OpenAI, 2026Vendor

This entry is educational. It is not legal, financial or professional advice, and it does not recommend any vendor. Figures attributed to vendors are their own claims.

Ask your own business. Get the true number in seconds.

Veld connects your accounting, sales and operations data to the AI tools you already use, and keeps it running, so every answer comes from your numbers.

Book a demo