Roothr Lab · Research

Small teams. Big operations. Human control.

Roothr Lab is the research arm of Root HR. We study how AI lets a small business run like a much larger one, without losing the person at the centre of every decision, and how compliance becomes the platform for that growth rather than the cost of it.

Read the plan Get involved

The thesis

AI should make people more powerful, not less necessary.

Most of what a small business does with its people is administration: pay them correctly, keep them legal, hire the next one, cover the rota, answer the question. It is necessary, it is unforgiving, and it falls on the same two or three people who also have to run the business.

AI can now do most of that work. The question is what it is for. We think the answer is opportunity: every hour handed back should become a better decision, a new hire, a contract the business could not have chased before, or simply a manager who goes home on time.

That only happens if the system is built deliberately. Requirements before models. A person directing and approving, with the assistant doing the work in between. And a compliance foundation strong enough that growth does not create risk.

Roothr Lab exists to work out how, to test it with real businesses, and to publish what we find.

How we build

Four rules for any AI we ship.

Requirements before models

Every piece of AI we build starts with three written answers: why is it needed, how does the whole system work end to end, and who stays in control. If we cannot write them down, we do not build it.

People direct, AI executes

The person names the outcome and approves the action. The assistant does the collecting, calculating, drafting and checking in between. Control never moves from the human to the machine.

Measure quality of life, not just output

A feature earns its place when it gives someone their evening back, removes a fear, or opens a door they did not have before. Hours saved is the floor, not the goal.

Compliance is the foundation of opportunity

A business that is confidently compliant can sponsor talent, bid for contracts, grow into new regions and sleep. We treat regulation as something to master, not something to survive.

The shift we are building for

The next powerful companies will be small.

AI lowers the number of people it takes to run an operation. That does not shrink ambition, it moves it. We expect the coming years to belong to SMEs and micro-companies with strong ideas and very few people, and they will need a backbone that keeps them compliant, efficient and connected.

Small teams run big operations

Five people with the right assistant can carry the payroll, compliance and people load that needed a department.

Every person gets more powerful

A manager who can ask for a shortlist, a report or a compliance check in one sentence makes decisions the size of a much larger company.

Resources go where they are needed

Rotas, budgets and hiring allocated on live data instead of habit. Less waste, fewer gaps, and the money kept inside the business.

Micro-companies become normal

Ideas that used to need funding to operate can now start with a handful of people and a set of agents. We expect a wave of them, and they will need a compliant backbone.

The system

HR software as the gateway.

The employee record is the one thing every business process touches: pay, hiring, rota, compliance, training, leaving. That makes the HR system the natural place for AI agents to connect, and the natural place to enforce who may see and do what.

People ask in plain English and approve every action
roothr.ai and partner agents do the work through one permissioned gateway
Regulators and government systems get evidenced reports, on time
Root HR stays the single source of truth throughout

Your teamasks and approves

AI agentsroothr.ai and partners

GatewayRoot HREmployee record · permissions · audit log · single source of truth

Regulators and governmentHMRC, UKVI, evidenced reports

Research programmes

Six questions we are working on.

Programme 1

Compliance intelligence

Turning UK employment, immigration and payroll rules into something a system can check continuously, so a business knows it is compliant every day rather than finding out at an audit.

Can sponsor duties, right to work rules and reporting windows be expressed as checkable rules that update when guidance changes?
What does a trustworthy "are we compliant?" answer look like, and what evidence must sit behind it?
How early can a business be warned before a breach, and what is the cheapest action that prevents it?

Programme 2

Decisions stay with people

Designing approval, explanation and audit so that a person stays genuinely in charge of payroll, hiring and compliance decisions, and never has to approve blindly.

What must an assistant show so that a ten second approval is an informed one?
Which decisions should never be automated, and how should a system refuse gracefully?
How do we log actions so that an auditor, a regulator or a tribunal can reconstruct exactly what happened and who decided?

Programme 3

Resource allocation for small teams

Helping a small business put people, hours and money where they create the most value, using the live rota, leave, cost and demand data already in Root HR.

Can a care home, a restaurant group or a warehouse predict staffing gaps a week out and fill them before they cost anything?
What is the real cost of overtime, agency cover and late hiring, and how do we make it visible in time to change it?
How do we recommend without nagging, and let the manager stay the manager?

Programme 4

The agent gateway

Root HR as the permissioned gateway through which other AI agents and services act on behalf of a business: accountants, recruiters, training providers, government systems and the business’s own agents.

How does an external agent prove who it is acting for, and what is it allowed to see and do?
What is the smallest, safest interface for an accounting agent, a recruitment agent or a UKVI reporting flow to plug into?
How do we keep the employee record as the single source of truth when many agents are reading and writing?

Programme 5

People to people opportunity

AI creates opportunity, but people close it. We are studying how small businesses find each other, share what works and turn patterns into repeatable playbooks.

Which operating patterns separate the small businesses that grow from the ones that stall, and can they be taught?
How do we connect an employer with a skill gap to a person or a partner who can fill it, without becoming a noisy marketplace?
What does a community of SME operators need from us: data, introductions, templates or something else?

Programme 6

Regulation ahead of time

Government rules on AI, employment and immigration are changing quickly. We want small businesses to meet each change already prepared, and to treat the change as a moment other businesses are not ready for.

How do we track upcoming UK regulation and translate it into concrete actions months before it lands?
What should an employer be able to prove about how AI was used in a hiring or pay decision?
Can readiness for a new rule be a competitive advantage we can measure?

The plan

What Roothr Lab will do, in order.

Dates are intentions, not promises. We will update this page as work ships and as what we learn changes the order.

Now · Q4 2026
Foundations
- Ship roothr.ai to the first Root HR teams with payroll, compliance and recruitment questions
- Publish the requirements template we use before building any AI capability
- Baseline study: how many hours a week a five person team spends on HR, payroll and compliance admin, and what it costs
Next · H1 2027
Continuous compliance and the gateway
- Continuous compliance checks for sponsor licence holders with a daily, evidenced status
- First external agent integrations through the gateway: accounting and recruitment
- Staffing gap prediction pilots in care, hospitality and logistics
- First Roothr Lab report: what changed for the early access teams
Later · H2 2027 onwards
Small teams, big operations
- Resource allocation recommendations across rota, budget and hiring
- Regulation radar: upcoming UK rules translated into dated actions for each customer
- SME operator community and shared playbooks
- Open research partnerships with universities and industry bodies

What we measure

Quality of life is the metric.

Hours returned

Admin hours per week per team member, before and after, and what those hours were spent on instead.

Compliance confidence

Days per month a business can prove it is compliant, and how far ahead of a breach it was warned.

Resource efficiency

Overtime, agency cover and unfilled shifts as a share of payroll, tracked month on month.

Opportunity created

Hires made, contracts bid for, licences obtained and regions entered that the team says it could not have managed before.

Get involved

Opportunity is made between people.

AI creates the opening. People walk through it together. Here is how to work with the lab.

Small and medium businesses

Join the early access group, tell us what eats your week, and help us measure what changes.

Get early access

Researchers and students

Work with us on compliance rules, approval and audit design, or resource allocation. We can share anonymised, consented data and real problems.

Write to us

Builders of agents and tools

If you are building an accounting, recruitment, training or compliance agent for UK businesses, talk to us about the gateway.

Talk about integration

People who want to build this

We are a small team with a big remit. If this page reads like your own thinking, we would like to meet you.

Join the lab