Neural · Industrial
A rounding error, multiplied by five hundred sites.
Consumable usage at one site is noise. Across a few hundred, it is a budget line nobody can explain. The data to explain it already exists in the dispensing records, the ERP and the purchase history, and reading it the way it would need to be read is nobody's full-time job.
Read this first. Neural is the agent architecture we run inside our own company. It has not yet been deployed inside a manufacturer, and you would be the first. Every capability below is labelled for exactly where it stands, and the labels separate what runs at CKS from what is in client beta from what is only designed.
Contents
- 01Where it goes wrongThree places, in the order they cost you
- 02One consumable, one weekThe whole loop, start to finish
- 03The workforceFive workflows and exactly where each one stands
- 04What we've actually doneStated precisely, including what we haven't
- 05DeploymentThe questions your IT group will ask first
- 06How it startsOne workflow, sixty days, read-only
[01]Where it goes wrong
Three places it costs you.
Usage nobody can account for
Gloves, blades, PPE and the expensive shared tool that never came back. Each one is trivial and none of them is trivial at scale. The variance shows up as a number in a budget review with no story attached to it.
The stockout costs more than the part
A line waiting on a consumable that costs a few dollars is the expensive version of a cheap problem. The signal that it was coming was in the consumption trend a week earlier.
Every site does it differently
The same process runs a slightly different way in each plant, so the corporate number is an aggregate of things that aren't quite comparable. Nobody trusts it enough to act on it.
[02]One consumable, one week
What it looks like across sites.
Written in the present tense for readability. This is a designed sequence, not an account of something that has happened at a customer.
- Mon
Dispensing volume for one consumable at one plant moves outside the band that plant normally runs, adjusted for shift pattern and headcount.
- Mon
The agent layer checks it against the other sites running the same process, and against that plant's own history, before treating it as real.
- Tue
It reaches the operations manager with both comparisons attached, not as an alert but as a question worth asking.
- Tue
Turns out to be a new crew and a training gap, not shrinkage. She marks it, and the system learns that shape.
- ·
The same read catches the reorder point that was about to be missed at two other sites.
It asks the question. A person decides whether it's a problem.
[03]The workforce
Five workflows, and where each one actually stands.
Every one has the same shape: a trigger, agent work, a human checkpoint, an artifact. The status under each name is the honest part, and it names where a workflow has actually run rather than how finished it feels.
Consumption anomaly
Trigger
Continuously, per site and per item.
Agent actions
Compares dispensing volume against that site's own baseline and against comparable sites, adjusted for shift pattern and headcount.
Human checkpoint
The operations manager marks it real or noise, and the system takes the correction.
Output
A short list of variances worth a question, with both comparisons attached.
Stockout forecast
Trigger
Nightly, across every site.
Agent actions
Projects consumption against on-hand and lead time, and surfaces the reorder points about to be missed rather than the ones already missed.
Human checkpoint
Supply chain confirms before anything is ordered.
Output
A ranked reorder list with the days of cover behind each line.
Tool accountability
Trigger
On check-out and check-in exceptions.
Agent actions
Tracks what left, what came back, and what has been out long enough to be worth a conversation.
Human checkpoint
The site supervisor follows up. The system never contacts an employee.
Output
An exception list by site, oldest first.
Executive brief
Trigger
Every morning, before anyone is at a desk.
Agent actions
Reads across the systems that hold your operation and writes what changed, what slipped, and what needs a decision today.
Human checkpoint
You reply in the thread; the work routes from there.
Output
One brief, in the order you'd have asked for it.
Compliance evidence
Trigger
Ahead of a safety audit or a customer requirement.
Agent actions
Assembles the issuance records showing what was stocked and drawn, which is the availability half of the evidence. It does not evidence use, fit or training, and it says so.
Human checkpoint
EHS reviews and signs the package.
Output
An evidence pack with its own gaps listed first.
[04]Standing
What we've actually done.
Stated precisely, because in this category precision is the differentiator.
We built the inventory system this reads
A connected vending and inventory platform deployed across 500+ industrial sites in manufacturing, mining and automotive. Four machine types, per-employee access limits, offline-capable devices in the field.
We've done the integrations
ERP integrations, real-time reporting and secure cloud data handling from physical devices, delivered as a plug-and-play install rather than a systems-integration project.
We run our own company on the agent layer
Neural plans, builds, reviews and ships at CKS every day. Demonstrable on a call.
The industrial workflows are designed, not deployed
Specified against the platform we built, which is why they talk about dispensing records and shift patterns. They have not run inside a manufacturer.
What we haven't done: run these agent workflows inside a manufacturer.
[05]Deployment
The questions IT will ask.
Answered here rather than in the third meeting.
- Where it runs
- Your infrastructure or a private tenant. Not a shared multi-tenant product.
- What it reads
- The systems you already run. Read access first, always.
- Who approves
- Actions you designate as consequential stop at a named person before they happen. You set that boundary; the platform enforces it rather than relying on discipline.
- What's recorded
- Each agent action is logged with its inputs, the sources it cited, what it did, and who approved it. The things that hold up in a review.
- Who owns it
- Your data and configuration are yours, and we'll put the exit terms in the contract rather than asking you to trust a web page.
[06]How it starts
One workflow. Sixty days.
Not a platform rollout. We pick the single workflow where it is costing you most, stand it up against your real systems, and you judge it on whether it caught things your people did not have time to.
Week 1
The workflow and the success metric are agreed in the first session, not discovered over a month. Read-only access to the systems it needs; your security review runs in parallel.
Weeks 2–7
It runs alongside your people on live work. Every output reviewed by your team, nothing acted on automatically, false positives tuned weekly.
Week 8
Measured against the metric we set in week 1. Keep it, change it, or walk away. Because it stayed read-only, walking away means revoking credentials and removing an integration rather than migrating off a platform.
- What it costs
- Fixed fee, quoted after the working session, so it is priced against your actual systems rather than a guess. No per-seat licence during the pilot.
- How it's judged
- One number, agreed before we start. Usually exceptions caught that your team didn't have hours to catch, measured against a baseline period.
- What we need
- Read-only access to two or three systems, one person who reviews the output, and an executive who reads the weekly.
- What you keep
- The configuration and everything it produced, whether or not you continue. Data handling and retention are in the agreement before access is granted.
A working session is ninety minutes with your operators, not a pitch. We map where the cost actually lands and tell you which parts of it are worth pointing an agent at, including the parts that are not.
Other sectors