Electronics and IoT

Can an agent change a setting on a customer's device without asking?

Your agents now reach devices in the field, the factory and the supply chain. LeafCraft sets what each one may do alone and what needs a person first.

One teamYour rulesYour memoryYour budget

Book a demo Experience now

LeafCraftto Karan Shah, Product operations lead · 10:42

Firmware setting change on customer devices

ItemFleet F-204 · 1,180 devices · power profileWhy youCustomer approval required by your ruleAgent saysApply. Resolves the overheating alerts from the last week.WaitingMoves to Deepa Nair, Head of device operations, in 2 hours

Your reason (required): "Customer confirmed the maintenance window."

ApproveSend backReject

Your reason is saved in the record, next to the agent's reason.

Decision request in Slack · example dataLeafCraft console

Trusted by

What's changing in electronics and IoT

Agents can now touch devices

An agent that changes a setting on a customer's device acts in the physical world. That needs a rule, not a hope.

Quality and warranty data in one loop

Field returns, test results and supplier claims feed each other. Agents move between them.

Supply planning on thin margins

A forecast agent that over-orders one component ties up cash for a quarter.

Two risks the layer removes

The agents you can't see, and the model that gets retired.

Agents you don't know about

In electronics and IoT, agents arrive through vendor updates and team pilots. LeafCraft finds all of them, gives each an owner, and brings it under your rules or removes it.

Inside software you buySwitched on by vendor updates

CRMERPHR systemService deskOffice suite

Built by your teamsPilots, scripts and side projects

LangGraphCrewAINotebooksAutomationsCopilot studios

LeafCraft connects to your ecosystem and finds every agent

One registerexample counts

23found

9no owner yet

11governed

3retired

How discovery works

A model that gets retired

An agent built straight on one model has to be rebuilt and re-approved when that model is retired or repriced. On LeafCraft you switch the model for that agent in config and keep everything else. Months and a budget, saved.

Agent tied to one model

The model is retired or repriced
Prompts, tests and sign-offs rebuilt
Months of work, a new approval cycle
The history stays with the old tool

MonthsA rebuild budget

Agent on LeafCraft

Pick another model for that agent
Switching is a config change
Test it against your own record first
Rules, limits, owner and memory stay

HoursNo rebuild

TimeHours to switch, not months to rebuild

CostNo rebuild budget, no new approval cycle

MemoryThe record carries over to the new model

Your choice of model

Why it matters now

In electronics and IoT, the agents arrived before the rules did.

Electronics and IoT companies run agents that read telemetry, plan supply and talk to devices in the field. The question is not whether an agent may act, but which actions it may take alone, which need a person, and whether the customer approved first. The layer makes that explicit per agent, and records every action it takes.

What the record answers hereEvery device action, who allowed it, and the quality and supply decisions around it.

How the layer works

Device management platforms

Quality systems

Planning notebooks

Engineering laptops

No shared owner · no shared limit · no shared record

LeafCraftone layer, on your side

One ownerper decision, AI or person

Your ruleson every run

Your memoryboth reasons, your database

Your budgetset before work runs

The core infrastructure

What LeafCraft runs for electronics and IoT.

One team

Quality and field issues go by impact to named people: the line lead, the quality manager, then the plant head.

How it works 02

Your rules

You decide which device actions an agent may take alone, which need the customer or a lead first, and which it may never take.

How it works 03

Your memory

Every device action, quality call and return decision is kept with the reasons of the agent and the person, in your environment.

How it works 04

Your budget

Every task and agent has a cost limit set before work runs, so a fault that keeps retrying does not run up a bill overnight.

How it works

Build and control a smart ecosystem

A smart ecosystem for your products, under your rules.

Your devices are in homes, factories and the field, and agents now act on them: reading telemetry, changing settings, pushing firmware, opening service tickets. LeafCraft lets you build that ecosystem and control it. Every agent that touches a device acts within a rule you wrote, with the user's approval where you choose, and every action is on a record that lasts the life of the product.

Devices in the fieldTelemetry, settings, firmware

Companion appsUsers and their approvals

Service and supportTickets, returns, repairs

Supply and qualityParts, tests, suppliers

LeafCraftyour rules · your memory · your budget

MQTT

Govern every device action

An agent may read a value, change a setting or push an update only within its rule. You choose which actions need the user to approve first.

Memory for the life of the product

Every action on a device, every service decision and every quality call, kept with the reason, from first boot to end of life.

Audit for safety and regulators

Show what changed on which devices, who allowed it and why, in minutes. Nothing erased after it is written.

Partners inside the ecosystem

Agents from component suppliers, service partners and platform vendors join under your rules, with their own login and limits.

Devices, events and platforms LeafCraft connects to

AI and people, one team

Your electronics and IoT team, with AI on the org chart.

Does the work

Decides

Checks

AI drafts. A person decides. Another agent checks.

The agent prepares the case. Your person makes the call and writes why. A separate audit agent checks the decision against your rules.

Same rules for every seatNo one checks their own workOne named ownerCost set before it runsBoth reasons recorded

No one checks their own workAgent or person. Maker and checker are never the same.

One named owner per decisionNever "the system".

Authority earned from the recordWidens with clean work. Narrows after mistakes.

People stay where they workSlack, Teams, WhatsApp (coming) or email. Their reply moves the work on.

What changes for you

Today

With LeafCraft

An agent changes device settings without a trail

Every device action has an owner, a rule and, where you choose, a customer approval

Quality decisions split across tools

One record across returns, tests and supplier claims

Planning agents retried all night

A failing step stops and asks a person

Model costs rise mid contract

Switch the model per agent in config

Silence is never a yesEstimated before it runsNothing erased after it's written

Record · Fleet F-204 setting changeKept in your database · unchanged since written

Export JSON / CSV

08:10:12

Device operations agent

Proposed a power profile change for 1,180 devices.

Why: Overheating alerts above threshold for 7 days.

08:10:13

Device operations agent

Passed to Karan Shah.

Why: Device changes need a person by rule.

09:02:45

Karan Shah

Approved.

Why: Customer confirmed the maintenance window.

09:02:46

Audit agent

Checked the decision against the rules. Passed.

Why: Maker and checker differ. Within cost limit.

Cost of this task2.40limit 6.00

OwnerKaran Shahnamed, never "the system"

ChecksPassedno one checked own work

Product screen · example dataLeafCraft console

LeafCraft sells no agents, no models and no compute.

NeutralGains nothing from your choice of agent or modelYoursRuns in your cloud, VPC or on-premOpenAny model, any cloud, MCP and A2AEqualBought, built and people, same rules

Why that matters

How it helps your people

Who it helps in electronics and IoT.

Leaders who ask first

CIOs, CTOs and IT headsVendors shipped agents and my teams built others. How do I govern all of them from one place?Read COOs and supply chain headsWhat happens when an agent changes a PO or a supplier order at 2am?Read Chief AI Officers and AI headsEvery new agent needs its own guardrails and tests. How do we scale past the first few?Read CISOs and security teamsHow many non-human identities do we have now, and what can each one reach?Read AI implementation teams and forward deployed engineersHow do I get an agent from demo to production without building the control layer myself?Read

Functions where it runs

Supply chain and procurementWhat stops an agent from changing a supplier order it should not?Read Customer serviceWhen a refund goes wrong, can we show who decided and why?Read IT operationsAgents can now grant access and change cloud settings. Who checks them?Read

Questions teams ask

Which device platforms can agents reach?

Agents can reach phones and connected devices through MQTT. Each action is one thing, like reading a value or changing a setting.

Can work start from our device events?

Yes. Agents pick up work when a message lands on Kafka, Amazon SQS, MQTT or Redis Streams, or on a schedule you set.

What stops a faulty agent from repeating a bad action?

A step that keeps failing is stopped and handed to a person, and outside agents that keep failing are cut off. One switch also pauses the whole operation until you resume.

Bring one electronics and IoT process.

We'll map it with you: which roles are AI and which are people, who decides, who checks, what it costs and what gets recorded.

Book a demo Experience now