How it works

Four things to understand, and none of them are technical.

You don't need to know how Leaf is built to know whether you can trust it. You need to know how work reaches your team, who is making the decisions, what happens when it's wrong, and what it remembers.

1

Work comes to your team one job at a time. Talking is optional.

Leaf sends the next job to the device in your hand. It can read it out through a headset and listen to the answer, or it can simply show it on the screen and wait for a tap and a scan. Same job, same rules, same record. The only thing that changes is how it is delivered.

Plenty of people do not want to talk to a computer all day, and plenty of tasks do not suit it. A packer at a bench with a screen in front of her is better off tapping. A picker with a pallet truck and both hands full is better off speaking. Leaf does not force a choice on anybody, and it is a per person setting, not a per site one.

When voice is used, the speech runs on the device itself. That makes it quick, it copes with a noisy building, and it keeps working when the warehouse wifi does not. Whichever mode you are in, the jobs already in hand carry on offline and catch up when the signal returns.

Voice and headset
Hands and eyes stay on the freight. Best for picking, putaway and counting.
Screen and scanner only
No headset, no talking, nothing listening. Read the job, scan, tap done.
A bit of both
Voice in the aisles, screens at the benches. Set per person, per task.
What this changes
Training goes from days of software instruction to a shift of shadowing, because there is one job on the screen instead of a menu of forty. Agency staff are productive on their first morning. And nobody is standing at a terminal typing up what they did an hour ago.
Leaf Offline, 6 jobs held
Putaway 3 of 14 Voice off
Take the pallet at dock four to 12-B-04. It is the second bay past the fire door.
Pallet L-44120 18 cases
FromDock 4
To12-B-04
Confirm byScanning the bay label
Bay is blocked. There is a cage in front of it.
Understood. Put it in 12-B-07 instead, same aisle, two along. I have flagged 12-B-04 for the shift lead.
Acted without asking. This job is set to act and notify.
Scan bay label
Talk
2

It uses judgement for the odd cases and arithmetic for the rest.

This is the honest bit. Most of what a warehouse does is not a judgement call, it's a calculation, and calculations should be exact and instant every single time. Leaf keeps the two jobs separate on purpose.

The routine work is exact

Shortest walk, best box, cheapest compliant carrier, who's free next. These are solved with established mathematics, the same methods the biggest operations in the world use. Nothing is invented, nothing is approximate, and the answer never varies with the weather.

The judgement is where it earns its keep

A truck with no paperwork. A supplier who sent the wrong thing. A cutoff you're about to miss. These are the moments that used to need your most experienced person. Leaf weighs them up, explains the trade-off, and either acts or asks, depending on what you've allowed.

Scanning never waits for thinking

When a picker scans a location, the confirmation is immediate, that path has no clever machinery in it at all. Your throughput doesn't depend on how busy the intelligence is, which is exactly how it should be.

3

You change how it behaves by telling it, in plain English.

No change request. No consultant on site for a fortnight. No configuration screen with four hundred checkboxes that only one person understands. You say what you want, Leaf shows you what it would have done differently over the last month, and then you switch it on.

You say
“For Acme's dropship orders, only ever use the carrier on their routing guide, and get the shipping notice out within an hour of the trailer leaving.”
Leaf answers
“Understood. I've tested that against the last 41,000 orders. It would have changed the carrier on 312 of them and cost $186 more in freight, but it would also have avoided nine late-notice chargebacks. Turn it on?”
You say yes. If it turns out wrong on Thursday, you say so, and it goes back to how it was, immediately, with nothing left half-changed.
Your customers' rules
Hand Leaf a retailer's routing guide and it becomes a set of rules the floor follows exactly.
Your contracts
Give it a client's rate card and it prices last month with it, so you can check before you bill.
Tested before it's live
Every change is rehearsed against your own history first. You see the difference before the floor does.
Undone in seconds
Going back is one instruction. There is no partial state to clean up afterwards.
4

Nothing is ever overwritten, so nothing is ever lost.

Leaf doesn't keep a picture of how the warehouse looks now and update it as things change. It keeps the whole story of what happened, in order, and works out the current picture from that. It sounds like a small distinction. It's the reason for almost everything good about living with it.

“Prove this shipment left on time”

The scan, the timestamp, the person, the trailer, the label. Not a summary written afterwards, the actual moment, still there.

“This batch is being recalled”

Every location it sat in, everything it was stacked with, every order it went out on. Minutes, from a standing start, at two in the morning.

“Your storage bill is wrong”

Leaf replays the month and shows what was really on hand each day, with the receipts and picks behind every figure. Disputes stop being archaeology.

“Why did it do that?”

Every decision was recorded with what it knew, what else it considered, and how confident it was. The explanation is looked up, not made up afterwards.

It's also how Leaf earns more responsibility. Because every decision and every human correction is on the record, you can see exactly how often it was overruled on a particular job, and let it off the leash when the number says it's ready, not when we say so.
For the technically curious

The stem and the veins.

Traditional warehouse software is one large program with modules bolted to the side of it. Leaf is shaped like a leaf. There is a stem, which is the small set of services that must never be wrong, and there are veins, which are sixteen AI agents, one for each part of the operation.

Every vein is genuinely an agent. It reads what is happening, reasons about it, calls on proven mathematics when a calculation is needed, and proposes what to do. The stem is what decides whether that proposal is allowed and what actually gets committed.

Receiving
Shipping
Putaway
Yard
Layout
Returns
Stock accuracy
Labour
Restocking
Quality
Order release
Assembly
Picking
Billing
Packing
Customs
Record
Dispatcher
Rulebook
Register
Simulator
Connectors
Conversation
Down the middle, the stem. Seven services with no AI in them. They hold the record, hand out the work, enforce the rules and connect the outside world.
Off it, sixteen veins. One AI agent per domain. Each one attaches only to the stem, never to another vein, which is why a new one can be grafted on without touching the rest.
The stem
Seven services. No AI anywhere inside them, on purpose. This is the part that has to be exact.
The record
Every fact, in the order it happened, written once and never changed. Your stock figures are calculated from it, not stored separately.
The dispatcher
Holds the pool of outstanding work and matches each job to whoever can do it next, person or machine.
The rulebook
Your policies, plus the rules nobody can override. Checked on every single instruction before it is allowed to happen.
The register
Who and what is available right now, what each can lift, reach and read, and how fast they actually work.
The simulator
A working model of your building, tuned by replaying your real history. Every proposed change is rehearsed here first.
The connectors
Your accounting system, stores, carriers, brokers, customers' trading files and robot fleets, translated in both directions.
The conversation
The one way people interact with Leaf. Answers about the past come from the record, never from an improvised guess.
The veins
Sixteen AI agents, each expert in one domain. None of them stores anything of its own and none of them calls another one directly. They watch the record, and they publish back to it.
Receiving
Putaway
Layout
Stock accuracy
Restocking
Order release
Picking
Packing
Shipping
Yard
Returns
Labour
Quality
Assembly
Billing
Customs
Why nothing is wired to anything

A vein never phones another vein. Picking simply announces that a line was picked, and whoever cares about that, layout, packing, stock accuracy, billing, hears it. Adding a seventeenth agent means adding a listener. It does not mean opening up the sixteen that already work.

The agents have proper tools

Route optimisation, box packing, wave planning, forecasting and staffing maths run as exact solvers the agent calls like a spreadsheet. The agent chooses which question to ask and why. The solver answers it the same way every time.

New abilities arrive quietly

A new agent is switched on in watching mode. It sees real work, records what it would have done, and changes nothing. Only once its answers have been scored against reality does it get to act, starting at the suggestion setting.

Where the AI is, and is not

Speech runs on the device in your hand. Judgement runs in the agents. Deciding what is permitted, and writing it down, is plain deterministic code. Nothing on the scan and confirm path waits on a model, which is why the floor never slows down.

Bring your own AI

Point whichever assistant you already use straight at your own data.

Leaf publishes a connection that AI assistants understand, using the open standard for it. That means you are not limited to asking Leaf questions inside Leaf. Connect the assistant your business already runs on, whether that is Claude, ChatGPT, Copilot or something your own developers built, and it can query the warehouse directly, in your words, with no export and no integration project.

The same door is open to your clients. A brand you store and ship for can connect their own assistant and ask about their own stock, their own orders and their own invoices, at two in the morning, without ringing your account manager or waiting for a report.

Nobody has to be given a login to your system, and nobody sees a row that is not theirs.

Your own team asks
“Pull last quarter's parcel margin by client and put it in the board pack.”
Your client asks their own assistant
“How much of our summer range is still sitting in the warehouse, and what did storage cost us last month?”
Your client's own software asks
“Hold every order for this batch and tell me which ones already shipped.”
Separated at the source
A connection is tied to one client. It is not a filter that somebody has to remember to apply. The data for everyone else is simply not reachable through that door.
Reading is safe by design
Outside assistants see a prepared view of the facts, not the raw internals, so nothing about your other clients or your own costs can be worked out from what comes back.
Writing goes through the rulebook
An outside assistant cannot change anything directly. It has to ask, through the same door your own floor uses, and the same rules can refuse it.
Every question is on the record
Who connected, what they asked and what they were shown, all recorded like any other event. If a client queries what they saw, you can show them.
Getting started

Nobody switches a warehouse over on a Sunday night.

So we don't ask you to. Leaf runs quietly alongside what you have now, making its decisions without acting on them, and you compare. When its answers are better than your current system's for long enough, you start letting it drive, one area at a time.

Weeks one and two
Watching
Leaf reads your stock, your clients and your contracts, and shadows the floor. It changes nothing.
Weeks three to six
Goods in
Receiving, putting away and counting move to voice. This is where teams stop wanting to go back.
The following month
Orders out
Picking, packing and shipping, with carrier costs benchmarked against your old system parcel by parcel.
When you're ready
The back office
Billing runs in parallel for a full cycle. The difference is either proof of parity or money you were losing.
The hardware is ordinary
Rugged handhelds, headsets, label printers. If you already have scanners, we'll tell you honestly whether they'll do.
Your data stays yours
Every client's stock and paperwork is walled off from every other client's, and each of them can connect their own tools to see only their own.
Robots when it suits you
Leaf hands work to machines the same way it hands it to people. Automating an aisle later doesn't mean buying another system.

Still sceptical? Good.

Come and try to catch it out. Bring the exception that breaks your current system and we'll run it live, then replay it and show you why it went that way.

Book a walkthrough