From customer orderto production floor.

ALT Steel manufactures roofing and building materials across multiple locations.

The systems challenge was to connect what happens commercially — customer orders, payments and sales — with what happens physically inside the business: steel coils, production jobs, machines, quality control and dispatch.

The resulting platform creates operational traceability between the customer transaction and the manufacturing process required to fulfil it.

Client
ALT Steel
Sector
Steel Manufacturing
Role
Systems Architecture & Implementation
Platform
Manufacturing Business Operating System
8
Production Machines Modelled
974.1m
Production Recorded
65%
OEE Visible in System
12
Steel Coils in Stock
13,576.4m
Coil Material Available
50,678kg
Tracked Coil Weight
257
Customers Visible
Operations Command Center — commercial, production, inventory and dispatch exceptions in one view.
The Challenge

A sale is only the beginning of a manufacturing workflow.

In a manufacturing business, recording an order does not mean the work is complete.

A customer order can create a requirement for physical production.

Management therefore needs visibility across both the commercial and physical sides of the business.

A disconnected system can answer one question: what did we sell?

  • Production Depends On

    • The correct product specification
    • Available raw material
    • An appropriate production machine
    • Production capacity
  • Fulfilment Depends On

    • Quality control
    • Dispatch or collection
    • Customer communication
  • Commercial State

    • Customer order
    • Payment status
    • Production authorisation
    • Order value

A manufacturing operating system has to reach past the transaction and describe the physical work behind it.

Operational Questions

What a manufacturing operating system must help answer.

  • 01

    What must we make?

  • 02

    What material will it consume?

  • 03

    Which machine should produce it?

  • 04

    Has it been produced?

  • 05

    Did it pass quality control?

  • 06

    Is it ready for the customer?

System Model

Connect the transaction to the physical work.

  1. Customer
  2. Order
  3. Payment
  4. Production Release
  5. Job
  6. Machine
  7. Steel Coil / Material
  8. Production
  9. Quality Control
  10. Dispatch / Collection

The system is designed around the relationship between commercial events and physical manufacturing events.

A customer order provides the demand. Payment can determine when production becomes authorised.

The production job converts that demand into work for the factory. The machine performs the work. The steel coil provides the physical material.

Quality control verifies the output. Dispatch completes the operational fulfilment process.

Management Control

Manage exceptions, not just records.

A command centre should help management understand where attention is required.

Instead of forcing users to inspect individual tables, the system surfaces operational states such as orders waiting for payment, jobs waiting for production, stock exceptions and dispatch readiness.

5
Awaiting Payment
1
Awaiting Production
18
Ready for Pickup
0
Production Delays
0
Low Stock Alerts
0
Critical Coils
12
Total Coils
2
Ready for Pickup — Dispatch View
1
Currently Loading

These are operational states recorded in the system at the time of capture. Different views can reflect different scopes or moments in the workflow, so the numbers are not forced to match.

  • Operations Command Center — exceptions across payment, production, stock and dispatch.
Workflow 01 — Order to Production

Production begins with authorised demand.

The production queue connects customer demand to factory activity.

A paid order that requires manufacturing can become a production job.

The important design principle is that the production floor should not operate as an isolated system. Commercial state determines what work is authorised to move into production.

  1. Customer Order
  2. Payment Status
  3. Production Requirement
  4. Create Job
  5. Job Queue
  6. Machine Assignment
  7. Production
  8. Completion
ORD-309048
Paid Order in Production View
US$72
Order Value
Ready
Status — Ready to Create Job

This is a single order visible in the production interface at the time of capture, shown as evidence of how commercial state feeds the production queue.

  • Production Queue — authorised orders can be converted into production work.
  • Production Floor — job cards provide visibility into manufacturing activity across machines.
Workflow 02 — Machine Operations

Machines become part of the operational data model.

Production capacity depends on physical machines.

The platform therefore models manufacturing equipment as operational entities rather than treating production as a generic status.

The system currently models eight production machines, including Machine 2 HQ, Machine 4 HQ, Machine 1 HQ, Machine 7 HQ, Machine 1 GWR and Machine 6 HQ.

The production view also exposes machine running state, job queue, meters produced, OEE, maintenance status and downtime.

  • Machine running state
  • Job queue
  • Meters produced
  • OEE
  • Maintenance status
  • Downtime
0 / 8
Machines Running at Capture
974.1m
Meters Produced
65%
OEE
0
Overdue Maintenance
0h
Downtime

These values represent the system state at the moment the screenshot was captured. They are not permanent performance metrics, and the platform is not presented as their cause.

  • Machine Operations — production equipment, running state and OEE visibility.
Inventory Intelligence

Inventory isn’t just “quantity on hand.”

Steel manufacturing requires inventory to reflect physical reality.

A steel coil is not simply an SKU with a quantity. Each physical coil can carry its own dimensional and consumption attributes.

The system therefore models individual steel coils as physical inventory assets, which creates traceability between manufacturing activity and physical material consumption.

Coil
  • Coil code
  • Colour
  • Thickness
  • Width
  • Weight
  • Initial length
  • Remaining length
  • Usage percentage
Material consumption
  1. Production Job
  2. Material Requirement
  3. Selected Coil
  4. Meters Consumed
  5. Remaining Coil Balance
  • Physical coil inventory with dimensional and consumption attributes.
Coil Intelligence

Turn physical inventory into production intelligence.

Once physical inventory is modelled correctly, the system can derive operational intelligence from it.

For example, remaining coil length can support estimates of available production capacity.

12
Coils in Stock
13,576.4m
Total Material Available
50,678kg
Total Weight
2,262
Maximum Sheets Remaining
151 days
Estimated Production Material Remaining
0
Critical Coil Alerts
0
Open Scrap Alerts

The displayed estimate of 151 days is based on the system’s recorded assumption of approximately 90m/day average production. It is a point-in-time estimate, not guaranteed future capacity.

  • Coil Intelligence — available meters, weight and estimated production capacity.
Workflow 04 — Quality Control

Production completion is not the same as accepted output.

Manufacturing workflows require quality controls after production.

The system includes inspection workflows so manufactured output can be evaluated rather than automatically considered complete.

NCR refers to a Non-Conformance Report. CAR refers to a Corrective Action Request.

1
Inspection
0
Passed
1
Failed
0
Open NCRs
0
Open CARs
0
Overdue Calibration

The single failed inspection is not an indication of overall production quality. It is evidence that the platform supports quality-control states and exceptions.

  • Quality Control — inspection results and manufacturing quality exceptions.
Workflow 05 — Dispatch

Production must eventually become fulfilment.

Different operational views can reflect different scopes or moments in the workflow.

The broader operational command view shows 18 orders as Ready for Pickup, while the dispatch view captured 2 ready for pickup and 1 currently loading.

  1. Production Complete
  2. Quality Accepted
  3. Ready for Dispatch
  4. Loading
  5. Collection / Delivery
18
Ready for Pickup — Command View
2
Ready for Pickup — Dispatch View
1
Currently Loading

These numbers are not forced to match. Normal operational differences between dashboard contexts are left visible.

Customer Liaison

Manufacturing still starts and ends with the customer.

Customer operations sit alongside manufacturing rather than outside it.

The goal is to allow commercial and customer-facing teams to understand fulfilment status without needing to manually reconstruct what is happening on the production floor.

The Customer Liaison area also supports inbound WhatsApp interactions.

257
Customers Visible
  • Customer Liaison — customer records and inbound communication.
Proof of Work

Inside the manufacturing system.

  • Operations Command Center — exceptions across payment, production, stock and dispatch.
  • Production — paid orders and factory job creation.
  • Machine Operations — production equipment and operating state.
  • Steel Coils — physical inventory attributes and remaining material.
  • Coil Intelligence — available meters, weight and estimated production capacity.
  • Quality Control — inspections, failures and corrective-action visibility.
  • Customer Liaison — customer records and inbound communication.
  • Production Floor — completed job cards across manufacturing machines.
Systems Thinking

The decisions behind the manufacturing model.

Decision 01

Payment can control production release

A customer order does not automatically become authorised production work. Commercial state and factory activity need to be connected.

Decision 02

Model the physical coil

Do not reduce manufacturing inventory to a generic quantity field. Width, thickness, weight, length and remaining material affect what the factory can actually produce.

Decision 03

Production consumes material

A completed job should have an operational relationship with the material used to produce it. This creates a foundation for material traceability and production intelligence.

Decision 04

Quality is a workflow

Manufacturing completion and quality acceptance are separate operational events. Inspection, non-conformance and corrective actions need explicit states.

Decision 05

Dashboards should surface exceptions

Management should be able to see what requires attention: orders waiting for payment, production waiting to start, material becoming critical, quality failures, maintenance issues, and fulfilment waiting for the customer.

What this project demonstrates.

  • 01Manufacturing process modelling
  • 02MES concepts
  • 03Order-to-production workflows
  • 04Physical inventory modelling
  • 05Material traceability
  • 06Machine operations
  • 07Production queues
  • 08Quality control
  • 09Maintenance visibility
  • 10OEE reporting
  • 11Dispatch workflows
  • 12Multi-branch operations
  • 13Customer operations
  • 14Operational dashboards
  • 15Executive intelligence

The system doesn’t treat manufacturing as a collection of forms. It models the physical events behind the business.

Responsibility

My role.

My role centred on understanding ALT Steel’s operational model and translating it into a connected system spanning customer demand, manufacturing, inventory and management visibility.

  • Business discovery
  • Manufacturing process mapping
  • Requirements definition
  • System architecture
  • Workflow design
  • Operational data modelling
  • Management-control design
  • Interface direction
  • AI-assisted implementation
  • Testing and iteration
Technology

Implementation tools.

  • React
  • TypeScript / JavaScript
  • Supabase
  • PostgreSQL
  • Tailwind
  • AI-assisted development
  • Web application architecture
Systems Insight
Model the physical reality first: coils, meters, machines and jobs — then let the dashboards emerge from those events.

I start with the operation, not the software.

ALT Steel demonstrates a different kind of systems problem from Island Group. Here the challenge is deeply physical. Customer demand becomes production work. Production work consumes real material on real machines. The manufactured output must pass through quality and fulfilment before the customer receives it. The software becomes useful when those events form one connected operational model.