STORIES · PRODUCTION
Scaling production starts where improvisation ends.
As an organization grows, so does the number of dependencies between sales, engineering, planning, production, logistics and cost control. Without a shared operating process, complexity starts growing faster than results.
Production rarely stops working overnight.
More often, workarounds, spreadsheets, local procedures and people's knowledge gradually begin to replace a coherent operating system.
The problem emerges when the organization is still producing, but answering simple questions becomes increasingly difficult: what is really happening, what does it cost and where is the constraint?
Digitalization makes sense when it brings order to the real flow of work, information and decisions.
4 STORIES
COMPLEXITY & SCALE
MTO · ERP · MES · WMS · BI · SCADA
How to scale complex production without scaling chaos with it
CONTEXT
Multi-stage production in a Make to Order environment, high product variability, demanding quality requirements, a complex supply chain and multiple dependencies between engineering, production and logistics.
PROBLEM
As the organization scaled, so did the number of local decisions, exceptions and dependencies. Optimizing individual workstations or departments was no longer enough because the result was created across the entire process chain.
CHANGE
Integration of operational and technology processes, development of ERP, CRM, BI, MES, WMS and SCADA environments, digitalization of information flows and automation of selected processes. In parallel, development of the production environment and technologies supporting new products.
MY ROLE
Responsibility for technology, operations and innovation, development of the production environment and digital transformation, from concept through implementation.
RESULT
The organization gained a more coherent view of production, costs and workflow. Technology became part of the operating process rather than a separate layer alongside it.
ADVANTAGE
Greater ability to scale complex production without proportionally scaling organizational complexity.
FROM THE PROJECT
The system would regularly slow down towards the end of every shift. The obvious answer seemed simple: more resources were needed.
Infrastructure, however, was not the problem.
Despite established processes, training and the requirement to report continuously in MES, supervisors would postpone data entry until the final minutes of the shift.
As a result, dozens of users would report hundreds of operations almost simultaneously, triggering thousands of dependencies and processing vast numbers of records.
Sometimes, before adding more computing power, it is worth checking what people are actually doing with the system.
TRANSFORMATION & CHANGE
SAP · ENTERPRISE · PMO · IT/OT · AUTOMATION
Why a new system is not enough if the organization still wants to work the old way
CONTEXT
A large-scale home appliance manufacturer with a complex supply chain, multiple operational processes and an enterprise-class systems environment.
PROBLEM
At this scale, every local change affects processes, data, integrations and the way other parts of the organization operate. Transformation therefore cannot mean implementing another set of independent technology solutions.
CHANGE
Building a digital transformation portfolio, developing project governance and PMO capabilities, advancing SAP and enterprise applications, integrating IT and OT, automating processes and establishing a more structured approach to technology-driven change.
MY ROLE
Responsibility for digital transformation and PMO, connecting business needs, technology architecture and project execution.
RESULT
Technology-driven changes became aligned with shared business priorities, dependencies and clear accountability for results.
ADVANTAGE
Not more technology. A greater ability to make change happen consistently.
FROM THE PROJECT
The new process worked differently from the one the organization had known for years.
The system was ready. The process had been designed. Users had been trained.
And yet new ideas began to emerge for how to continue working the old way inside the new environment.
The organization was capable of spending time, energy and budget not on changing how it operated, but on recreating the old way of working inside new technology.
Why?
Because the new was unfamiliar. And the unfamiliar is easily perceived as a threat.
The hardest part of transformation is rarely technology. More often, it is accepting that after the change, you really do have to work differently.
IN THIS PROJECT
- Digital Transformation
- SAP
- PMO
- Enterprise Applications
- IT/OT
- Automation
- Governance
PLANNING AND COMPETENCE
MTO · APS · MES · ERP · WMS · HR
How to stop planning production around machines and start planning it around people too
CONTEXT
Make to Order production operating across three shifts with a broad product mix differing significantly in complexity, risk and margin. Scheduling answered whether a machine was available, but not whether a particular team should execute a particular order.
PROBLEM
Complex and expensive products were assigned wherever capacity happened to be available. The consequences appeared later as defects, rework, exceeded lead times and lost margin. Quality data existed, but it did not feed back into production planning.
CHANGE
Competence became a planning resource alongside machines and materials. MES, ERP and WMS data was used to create capability profiles for teams, workstations and machines, while APS began allocating orders based on complexity, risk and demonstrated competence.
MY ROLE
Concept of the model, data-flow architecture across MES, ERP, WMS, APS and HR, and translating production events into reliable planning measures. From operating principles through implementation and operation of the model.
RESULT
Defects and rework decreased for the highest-value products, lead-time variability was reduced and manufacturing economics improved without expanding the machine park or headcount. Competence stopped being a manager's opinion and became a measurable organisational resource.
ADVANTAGE
The ability to deliberately place quality risk where it costs the least and competence where it protects margin the most.
FROM THE PROJECT
Simple displays at workstations showed plan, actual performance and the effect on bonuses.
The unexpected result was that employees began actively asking for training because they could see the direct relationship between competence, assignment complexity and pay.
People rarely work below their ability. More often, they work exactly the way they are measured.
IN THIS PROJECT
- Strategy
- Manufacturing
- APS
- MES
- ERP
- WMS
- BI
- HR
- Digital Signage
- Competence Management
- Cost Optimisation
AI GOVERNANCE
AI GOVERNANCE · EU AI ACT · NIS2 · GDPR · HUMAN OVERSIGHT
How to take AI from pilot to production without losing control
CONTEXT
A Make-to-Order manufacturer operating across several locations, with an international contract portfolio and a place in the global supply chain. Part of the technical documentation belonged to customers and was subject to contractual confidentiality and processing requirements. Within one year, seven AI initiatives emerged across the organization, from complaint analysis and technical documentation to quoting, maintenance planning and delivery-date prediction. Most of them were created bottom-up, outside the formal project structure.
PROBLEM
All the pilots worked. None reached production. The problem was not technological. There was no clear answer to questions about responsibility for an incorrect recommendation, whether the data could legally and contractually be used, where it was processed, how customer documentation was protected, or what level of human oversight was required. The EU AI Act, GDPR, cybersecurity requirements and obligations flowing through the international supply chain had different owners. No one, however, had the mandate to make one decision: deploy it or consciously decide not to. The pilots did not stop because they failed. They stopped because no decision could be made.
CHANGE
An AI Governance model was created that turned the question “are we allowed to do this?” into a repeatable decision-making process. We started by inventorying every AI use case, including tools already being used informally. A register was created with risk classification based on the logic of the EU AI Act, a business owner accountable for the outcome, and rules governing customer-provided data and intellectual property. Levels of human oversight were defined: where the model recommends, where it can act conditionally, and where the final decision must remain entirely with a human. Decisions were made by a three-person team representing business, risk and technology. Every request had to receive a decision within 14 days. The model also included monitoring of AI systems operating in production, logging of recommendations and periodic verification that each solution continued to do what it had originally been approved to do.
MY ROLE
Responsibility for designing and implementing the AI Governance model: decision structure, use-case register, risk classification, data-processing rules and model oversight. Translating regulation, contractual obligations and technology risk into criteria that allowed a decision to be made in 14 days rather than 14 months. Leading the model during its initial operating period and transferring ownership to the organization.
RESULT
Three of the seven initiatives moved into production. Two were deliberately closed. Two returned for further work with a specific list of gaps. The path from idea to decision went from “unknown” to two weeks. Tools previously used informally were brought under control or replaced with solutions meeting the organization's requirements. The organization gained an up-to-date register of AI use cases with risk classification, ownership and data-processing rules, ready not only for the day of an audit, but every day.
ADVANTAGE
The ability to deploy AI at the pace required by a demanding international supply chain. Risk stopped being an argument for inaction and became a decision parameter. Mature AI governance stopped being only a compliance cost. It also became an argument for credibility with customers.
FROM THE PROJECT
The technologist who had independently built an assistant for technical documentation arrived at the first meeting convinced that he would have to shut it down.
He did not. His solution became the first entry in the register and the first case taken through the entire path: risk classification, organization of data sources, migration of processing into a controlled environment, appropriate access controls and assignment of an accountable owner.
Three months later, the solution was being used by technology, quality and maintenance teams. Its creator began training other teams.
During the following six months, eleven new bottom-up proposals were submitted to the register. Previously, there had been none. Because there had been nowhere to submit them.
Shadow AI does not come from people being careless. It comes from nobody building them a door, so they climbed in through the window.
IN THIS PROJECT
- Strategy
- AI Governance
- EU AI Act
- Human Oversight
- NIS2
- GDPR
- Risk Management
- Shadow AI
- Intellectual Property Protection
- Supply Chain
- Compliance
- Change Management
COMMON DENOMINATOR
Different technologies. One organizational operating system.
ERP, MES, WMS, automation and analytics solve different problems.
But only when they work together with processes and human accountability do they begin to build an organization's ability to scale.
Digitalizing production is not about replacing the analogue world with a digital one. It is about designing a better way to operate.
EXPLORE OTHER STORIES
Technology creates value when it changes how the organization operates.
If you see a problem, an opportunity or a change that cannot be addressed with another standard project, let's talk about the result you want to achieve.