Skip to main content
← All case studies
Case Study — Company-Wide Transformation

From Disconnected Operations to an Intelligent, Connected Enterprise

How a specialty-materials company is connecting sales, procurement, piece-level inventory, production, machines, quality, dispatch and commercial intelligence through one operating system.

Multi-stage implementation currently in progressActive transformation architectureReal anonymised engagement

This is a real anonymised engagement. The client identity and confidential business information have been withheld. Quantified outcomes will be published only after implementation data is available and client approval is received.

  • Enterprise Operations
  • ERP Value Realisation
  • Piece-Level Traceability
  • Machine Connectivity
  • AI Command Center
At a glance

The engagement in summary.

Industry
Specialty steel distribution and processing
Business model
High-value material stocking, processing and customer-specific fulfilment
Transformation scope
Sales → Procurement → Inventory → Production → Machines → Quality → Dispatch → Customer
Primary challenge
Business transactions, material movement and physical execution were not controlled through one connected system
Solution
Enterprise Operations System and AI Command Center
Transformation status
Multi-stage implementation in progress
Strategic objective
Build a software-directed, traceable and intelligent industrial service operation
Business context

High-value materials require more than quantity-level inventory.

The company procures, stocks, processes and supplies specialty materials across different grades, forms, dimensions, heat numbers, metallurgical routes, quality specifications, customer applications and certification requirements.

Two physical pieces may appear similar while differing materially in grade, heat, dimensions, quality route, properties, certification, age and customer suitability.

Quality failure

Material loss

Production delay

Certification failure

Margin leakage

Customer risk

The challenge

The ERP recorded transactions, but physical execution remained fragmented.

A sales order could be recorded in the enterprise system, but many decisions that followed depended on calls, spreadsheets, paper records and individual experience.

Sales, procurement, stores, production, quality, dispatch and management each held part of the required information, but the company did not yet have one closed-loop operating system connecting them.

Is technically suitable stock available?

Which exact physical piece should fulfil the order?

Where is that piece located?

Is it reserved or under quality hold?

What are its grade, heat, dimensions and age?

Which material choice provides the best yield?

Which machine should process it?

How should it be cut?

When can the order be completed?

Has the correct material reached the machine?

Is the machine executing the authorised instruction?

What finished pieces and remnants were created?

What is the actual order profitability?

What can sales confidently promise the customer?

Transformation architecture

One intelligent operating layer across the enterprise.

  1. Customer Demand
  2. Commercial Validation
  3. Inventory Allocation
  4. Procurement
  5. Production Planning
  6. Machine Execution
  7. Quality
  8. Finished Goods
  9. Dispatch
  10. Customer Visibility

In this engagement, the ERP remains the financial and transactional system of record.

Xylm is building the management intelligence, orchestration and execution layer around it to determine, direct and verify how physical fulfilment occurs.

The ERP records the order. The connected operating layer directs and verifies how it is fulfilled.

Layer 1

Commercial and demand intelligence

Sales and order intelligence

  • Grade
  • Form and dimensions
  • Quality route
  • Heat-treatment condition
  • Certification
  • Quantity
  • Delivery requirement
  • Available stock
  • Suitable alternatives
  • Production lead time
  • Commercial exposure
  • Expected margin

Sales visibility can include available stock, reserved stock, work-in-progress, incoming purchases, material under quality hold, stock available after processing, expected completion and dispatch status.

Pricing and margin protection

  • Historical material cost
  • Current replacement cost
  • Foreign exchange
  • Freight
  • Energy and processing exposure
  • Material yield
  • Scrap value
  • Remnant value
  • Inventory age
  • Customer terms
  • Required margin

The commercial decision becomes connected to the physical and economic consequences of fulfilling the order.

Procurement intelligence

  • Technical compliance
  • Price
  • Delivery timeline
  • Payment terms
  • Freight
  • Tax
  • Inspection requirements
  • Historical supplier performance
  • Quality performance
  • Total landed cost

Requirement → Eligible Suppliers or Mills → Enquiry → Quotations → Comparison → Approval → Purchase Order.

Layer 2

Material identity, quality and inventory control

Digital identity for every physical piece

  • Grade
  • Form
  • Heat and batch
  • Supplier and mill
  • Quality route
  • Chemistry
  • Mechanical properties
  • Certificate
  • Dimensions
  • Actual weight
  • Storage location
  • Cost and replacement cost
  • Age
  • Reservation status
  • Quality status
  • Movement history

Dimensions are tracked at the physical-piece level rather than being embedded only inside a generic SKU.

Inward quality and certification

  • Purchase order
  • Supplier
  • Heat number
  • Test certificate
  • Chemical requirements
  • Mechanical requirements
  • Inspection status
  • Accepted quantity
  • Rejected or held quantity
  • Storage location

Materials failing validation can be placed under digital quality hold and prevented from allocation.

Piece-level warehouse control

  • Receipt, inspection and put-away
  • Reservation and picking
  • Production transfer and processing
  • Remnant return
  • Finished-goods storage and dispatch
  • Exact piece location
  • Available and reserved quantity
  • Material age, slow-moving and dead stock
  • Misplaced material and usable remnants
Layer 3

Optimisation, planning and machine execution

Intelligent material allocation

  • Grade compatibility
  • Heat and certification
  • Quality route
  • Dimensions and allowance
  • Location and reservation
  • Inventory age
  • Cutting loss and remnant usability
  • Material and replacement cost
  • Future demand and competing orders

The objective is not only to satisfy the current order. It is to protect total material yield, future availability and margin.

Cutting and yield optimisation

  • Finished output
  • Kerf loss
  • Scrap
  • Remnant size and future usability
  • Processing time
  • Profitability

An optimisation capability being designed and implemented — not a completed quantified result. The target recommendation meets customer requirements, minimises avoidable loss, preserves useful remnants, considers future demand and protects margin.

Production planning and machine-load optimisation

  • Order priority
  • Material piece
  • Machine eligibility
  • Processing timing
  • Cutting instructions
  • Job grouping
  • Changeover reduction
  • Machine load
  • Customer-delivery risk

Material-verified machine release

  • Material ID
  • Grade
  • Heat
  • Dimensions
  • Quality status
  • Customer reservation
  • Work order
  • Machine eligibility

Before: "The operator believes this is the correct material." Target state: "The system has verified that this is the material approved for the job." Safety-critical controls remain with machine PLCs, local safety systems and authorised personnel.

Two-way machine connectivity

  • Instructions sent: work order, job identity, material identity, quantity, cut dimensions, tolerance, approved parameters, job sequence, priority, operator authorisation
  • Information returned: job start and completion, actual production quantity, cycle and cutting time, feed and blade parameters, machine state, idle and stoppage time, alarms, deviations, blade utilisation, production history

Machine-connectivity architecture being implemented and validated. Not every machine is already connected.

Layer 4

Traceability, fulfilment and customer visibility

  1. Parent Raw Material
  2. Finished Customer Pieces
  3. Usable Remnants
  4. Scrap
  5. Process Loss
  6. Quality Results
  7. Machine Parameters
  8. Sales Order
  9. Dispatch
  10. Customer

Complete material genealogy

  • Grade
  • Heat
  • Supplier
  • Mill certificate
  • Quality records
  • Processing history
  • Authorisation
  • Customer destination

Finished goods and dispatch control

  • Quality release
  • Labelling
  • Packing
  • Storage
  • Dispatch planning
  • Vehicle loading
  • Shipment
  • Proof of delivery

Customer visibility

  • Order status
  • Planned completion
  • Shipment details
  • Weight
  • Invoice
  • Test certificate
  • Quality documentation
  • Historical orders

Potential customer-portal capabilities. Each is delivered according to actual implementation status; not every capability is live today.

Layer 5

Enterprise command centre and AI-supported decisions

Enterprise Command Center

  • Sales orders and order fulfilment
  • Inventory and procurement
  • Production and machine performance
  • Quality and dispatch
  • Customer commitments
  • Working capital, pricing and margin

Leadership exceptions

  • Orders at risk
  • Stock shortages
  • Ageing inventory and slow-moving grades
  • Allocation exceptions
  • Production bottlenecks and machine downtime
  • Yield loss and margin leakage
  • Delayed dispatch
  • Supplier underperformance
  • Customer-service issues

Deterministic controls

  • Material identity
  • Grade validation
  • Authorisation
  • Quality compliance
  • Traceability
  • Machine safety

Target AI and optimisation applications

  • Material-allocation recommendations
  • Cutting optimisation
  • Production sequencing and machine-load balancing
  • Completion-time prediction
  • Demand forecasting
  • Procurement recommendations
  • Slow-moving inventory prediction
  • Price and margin alerts
  • Machine anomaly detection and predictive maintenance
  • Exception prioritisation
  • Natural-language management queries

Target applications. They are not all live today. Example management questions: which orders are most likely to be delayed? Which material should be allocated to this order? Which grades are ageing without sufficient demand? Where was material yield lost? Which orders are below the approved margin? What should be produced next to protect delivery and utilisation?

Current and target state

Where the operation is today, and where it is going.

Current-state challenges
  • Fragmented departmental information
  • Material selection dependent on individual knowledge
  • Risk of incorrect material processing
  • Cutting decisions not optimising total material value
  • Manual physical-inventory verification
  • Machine activity disconnected from the enterprise system
  • Limited continuous traceability of finished pieces and remnants
  • Sales without fully verified availability and completion information
  • Profitability affected by hidden yield and pricing factors
  • Management time consumed by routine follow-up
Target state
  • Every material piece has a digital identity
  • Sales sees verified stock and completion information
  • Procurement connects technical and commercial requirements
  • Allocation considers suitability, yield and margin
  • Production is planned against inventory and capacity
  • Machines receive authorised instructions
  • Incorrect material prevents routine release
  • Actual machine execution is captured
  • Finished goods, remnants and scrap remain traceable
  • Customer documents remain digitally connected
  • Management works through prioritised exceptions
  • AI supports decisions using live enterprise data
Transformation status

Architecture and implementation scope are the current proof.

This is an ongoing, multi-stage transformation.

No percentage improvements for yield, inventory accuracy, machine utilisation, delivery performance or profitability are shown, because those results have not yet been measured after implementation.

  • Order-to-machine orchestration
  • Piece-level inventory
  • Digital material genealogy
  • Intelligent allocation and cutting design
  • Material-verified execution
  • Machine-connectivity architecture
  • Real-time enterprise visibility
  • Commercial and pricing intelligence
  • AI-supported exception-management design
Measures being established

The operating measures being established

  • Material utilisation
  • Scrap and unusable-remnant percentage
  • Wrong-material incidents
  • Inventory accuracy
  • Traceability coverage
  • Material-picking time
  • Order-to-production-release time
  • Production-plan adherence
  • Machine utilisation
  • Unplanned downtime
  • Actual versus standard cutting time
  • On-time delivery
  • Quote-response time
  • Inventory ageing
  • Actual versus quoted margin
  • Manual interventions per order

These are the measures the operating system is being designed to establish. They are not presented as completed results.

Strategic outcome

From people-dependent processing to a software-directed enterprise.

This is not simply an ERP, WMS, MES, machine-integration or dashboard project.

It is the creation of a closed-loop enterprise operating model in which every customer order becomes an optimised production decision, every material movement is digitally verified, every machine action is captured, every finished piece remains traceable and every operating decision is connected to its commercial consequence.

Order-to-machine orchestration

Customer demand converted into verified production instructions.

Piece-level control

Every raw material, finished piece and usable remnant digitally identified.

Material-verified execution

Incorrect material prevented from routine production release.

Complete enterprise traceability

From supplier and heat number to finished product and customer.

AI-supported decisions

Allocation, planning, pricing and exceptions evaluated using live enterprise data.

Where this model applies

Relevant for traceability-intensive, high-value and make-to-order operations.

  • Specialty metals and materials
  • Steel and alloy distribution
  • Industrial service centres
  • Precision engineering
  • Aerospace and defence components
  • Automotive components
  • Medical and regulated components
  • Engineered-to-order products
  • High-value batch or piece-level inventory
  • Operations requiring machine and material genealogy

These are potential application areas for the operating model. They are not presented as additional client references.

Are your transactions digital — but physical execution still people-dependent?