Work preparation is overloaded
Revisions, bills of materials, work instructions and exceptions take more time than planned. The shop floor waits for clarity.
Make visible where your operation gets stuck. Then build the right AI step.
A practical scan of a process, department or operating model in your production environment. In two to four weeks, you know which AI step creates value first.
Getting AI right starts in the operation itself: revisions, quality files, ERP actions, engineering handovers and knowledge held by a few people. The Opportunity Scan makes visible what good work looks like, where it gets stuck and which step is executable.
30 minutes with a BPC partner. No preparation needed.
After the scan, you can decide
01
Where to start
Work preparation, quality, service, planning or engineering handover
02
What to tackle
An agent, workflow or process improvement that fits your stack
03
Who is needed
The roles and decisions needed to move from analysis to execution
Sound familiar?
Production keeps running. The work around it keeps getting heavier.
Many manufacturers have AI on the agenda. The real opportunities often sit in work preparation, engineering, quality and handover to the floor.
Knowledge sits with the same few people
Senior employees become the helpdesk. That slows new colleagues down and makes the organisation vulnerable.
ERP is the system, workarounds are the reality
The formal process looks right on paper. In practice, Excel sheets, email routes and manual checks appear everywhere.
Quality needs more evidence
Files, certificates and checks need to be complete. A lot of time goes into searching, collecting and verifying.
What good looks like
You do not automate a process diagram. You make the best way of working transferable.
In manufacturing, a lot of value sits in the judgement of experienced people: when a revision needs attention, when a file is truly complete, which workaround creates risk. The scan brings that logic to the surface.
01
Follow the best way of working
We observe people who know how work preparation, quality, planning or engineering handover really works when it is done well.
02
Capture the exceptions
We make explicit which signals, checks and exceptions currently live in people's heads instead of systems or work instructions.
03
Build on decision logic
Only then do we determine which agent, workflow or process measure fits, including where human oversight remains necessary.
What you get
You see where AI frees capacity because it becomes clear what good execution requires.
The scan helps operations, production, engineering and IT choose the same priority. Starting from how the work really flows.
Discuss your operationTarget lead-time reduction
We show where waiting, searching, rework or repeated alignment extends lead time.
More grip on knowledge and documentation
We determine where a knowledge agent, documentation agent or control agent can have the most impact.
Better collaboration between floor and office
We reconstruct where handovers between engineering, planning, work preparation and production slow down.
A build candidate that fits your stack
We test what is possible within SAP, Microsoft Dynamics, PDM, SharePoint, Volentis or private AI.
How we work
We extract the logic from the daily work around production.
No abstract AI workshop. We look at how experienced people work, which choices they make and where information, systems and decisions touch.
Stap 1
Define the scope
Together we choose a process, department or broader operation. For example work preparation, quality, service, planning or engineering handover.
Stap 2
Talk to key roles
We speak with the people who carry the work every day: operations, team leads, work preparers, quality, planners and IT. We ask why they do one thing instead of another.
Stap 3
Look over the shoulder
We observe how actions in systems happen. The employee uses their own system. We capture what slows the work down, makes it vulnerable or depends on personal judgement.
Stap 4
Translate patterns into opportunities
We combine observations, interviews and documents into a concrete view of decision logic, agent candidates, automation and process changes.
Beyond the agent
An agent in production touches data, cost, quality and responsibility.
In manufacturing, AI quickly touches ERP, PDM, engineering IP, quality files and handover to the floor. That is why the scan already looks at the conditions that decide whether building is wise.
ERP, PDM and engineering data
Which sources are needed, where may information go and which knowledge or IP must stay inside your own environment?
Quality and compliance
We look at which controls, evidence, traceability and human oversight are needed before AI takes a role in the process.
Cost at scale
If planners, work preparers or service teams use the solution every day, inference volume and platform choice become part of the business case.
Ownership in the operation
We make visible who decides, who manages, who checks output and how exceptions or errors are followed up.
Choose your entry point
Start with one process or bring the whole operation into view.
The right scope depends on urgency. Sometimes you need quick proof in work preparation. Sometimes you need to design the role split between people, systems and agents more broadly.
Narrow and fast
One process
Duration
about 2 weeks
Indication
€3,500 - €4,500
Best for: work preparation, quality control, service intake or documentation where the pain is clear.
Outcome: one executable improvement with clear impact on time, quality or capacity.
Most chosen
One department
Duration
about 3 weeks
Indication
€6,000 - €8,500
Best for: operations, engineering support, quality or planning where processes touch each other.
Outcome: a set of opportunities and a first agent or process build candidate.
Strategic
Operating model
Duration
about 4 weeks
Indication
€10,000 - €15,000
Best for: leadership teams that want to understand how AI changes capacity, roles and decision making.
Outcome: a roadmap for human-agent collaboration in your production environment.
What you get
A plan that fits the reality of your factory, systems and people.
We do more than map opportunities. We show what you can build, what it requires and where the first business value sits.
Process reconstruction with decision logic
A concrete view of steps, handovers, systems, exceptions, recurring manual work and the choices experienced people make along the way.
Opportunity backlog with earning potential
A handful of concrete opportunities, prioritised on lead time, quality, capacity, feasibility and dependencies in IT and data, each with an indication of expected return and the effort required. You see what it returns, not only what is possible.
Why it has not happened yet
For each opportunity we name what actually held the step back (knowledge, ownership, systems or shop-floor data) and what it takes to remove that blocker.
Agent candidates with human boundaries
Proposals for knowledge agents, documentation agents, control agents or workflow support, including what the agent may do and where human judgement remains necessary.
Decision plan for the first build
A 90-day plan with roles, required input, risks and the first go/no-go moment.
Why BPC
We understand that AI in manufacturing only works when the process is right.
BPC works where operations, systems, transformation and AI meet. That is exactly where value appears in production.
Process knowledge from practice
We look at real actions, handovers, exceptions and judgement. Not only at the formal process.
From scan to build
The scan can lead directly to an agent, workflow, operating-model change or private AI solution.
Experience with industrial stacks
We build on existing systems such as ERP, PDM, SharePoint, Microsoft, Volentis or private AI.
FAQ
What manufacturers usually ask
Can we start small?
Yes. A scan on one process is often the best way to quickly show where AI or process improvement creates value.
Do you need access to ERP or PDM?
No. During the scan we look over the shoulder. The employee uses their own system. We only use additional observation methods with explicit permission and within clear agreements on scope, privacy and security.
Is this mainly for IT or operations?
For both. The buyer is often in operations or leadership, but IT needs to be involved early enough to assess feasibility, security and stack choices.
What can be built afterwards?
For example a knowledge agent, documentation agent, work-preparation agent, quality-check support or a broader operating-model change.
The next step
Where does your operation lose the most time today?
In 30 minutes, we determine where pressure sits in your operation and which scan variant makes sense. No preparation needed.
A practical conversation with a BPC partner. Focused on your operation.