Industry Manufacturing

AI in manufacturing that knows your machines.

AI consulting for machine builders: read specifications, find service knowledge, prepare documentation, connected to ERP, PLM and DMS.

Free intro callDiscuss your project
Two technicians in hard hats check a program at the control panel of a CNC machineAn older mechanic grinds a workpiece on the bench grinder in his workshopA machine shop with lathes, tool carts and chain hoists
  • wstnd
  • klassy
  • selectry.io
  • Gastro Möbel Bölme
  • ai-think
  • Revotion

Sound familiar?

The knowledge behind every quote and service call is in the building. Just not where it is needed. It sits in specifications, old quotes, service reports, manuals and the heads of a few colleagues.

  • Sales

    Quotes from specifications take too long.

    A sales engineer checks a thick specification against past projects by hand.

  • Service

    Service knowledge depends on people.

    The technician is on site and first has to reach the one colleague who has seen the fault before.

  • Technical writing

    Documentation eats engineering time.

    Manuals and variant documents pile up at the end of a project.

  • Engineering

    Experience retires.

    A long-serving design engineer leaves, and nobody quite knows why something was built that way.

The cost of waitingExperienced people leave every year, and with them knowledge that is written down nowhere. The later you capture it, the less is left.

The situation in figures

Where the industry stands today.

What associations and studies say about AI in the industry. Every figure with its source.

  • 89%

    Leadership is convinced.

    Almost all executives in machinery and plant engineering consider generative AI important for profitability. They see the most potential in sales, engineering and after sales.

    VDMA and Strategy&, GenAI in machinery, 2025 (German)

  • 71%

    Knowledge sits in a few heads.

    Seven in ten machinery builders cannot find enough suitable engineers. What the experienced people know has to become usable for everyone before they retire.

    VDMA press release, 6 January 2026 (German)

  • 20%

    Service carries the revenue.

    Around a fifth of revenue comes from service and customer support. Every hour a technician spends searching is missing exactly there.

    VDMA, service and customer support (German)

  • 2027

    Documentation goes digital.

    The EU Machinery Regulation applies from 20 January 2027. Instructions and declarations of conformity may and should be provided digitally.

    European Commission, Machinery Regulation

Where AI takes work off

What changes at your desks.

Not experiments, but processes that cost you time every day. Each one works in the systems you already have.

The right quote has been written before. Nobody can find it.

Today

Specification line 3page 14 of 82

4.2 Press requirements

The supplier delivers a servo press with controls, guarding and documentation.

Press force min. 4,000 kN

All values apply in continuous operation, three shifts on five days.

Cycle time max. 2.1 s at 600 mm stroke

Acceptance takes place at the customer's plant with both parties present.

Acceptance per EN ISO 16092

Payment terms see section 9.

  • Specifications are read by hand, page by page.
  • Requirements are retyped into Excel.
  • Earlier quotes are scattered across the shared drive.

With AI

Requirement list142 items
  1. Press force 4,000 kNStandard PX-4000met
  2. Cycle time 2.1 sData sheet 2.3 squestion
  3. Similar projectLine 1, 202392% similar

Quote ready as a draft

  • The AI marks every requirement in the specification.
  • Matching assemblies and comparable quotes are suggested.
  • The costing is prepared, your sales team checks and decides.

Faster to the quote, built on what you have already built.

Discuss this caseSee the use case

The answer exists. Just not where the technician is standing.

Today

Team chatService
  1. M. Keller, field service

    Anyone seen E-217 on the VX-400 before? Customer is waiting.

  2. S. Brandt, hotline

    Ask Juergen, but he is on holiday until Friday.

  • Questions go to colleagues who are on the road themselves.
  • Manuals and service reports sit in different systems.
  • In doubt, individual experience decides.

With AI

Service assistantCase 24-1187
  1. M. Keller, field service

    VX-400, SN 40218: error E-217. What do I check first?

  2. Check filter F-3 for clogging, then pump KP-12. In 3 similar cases F-3 was the cause.

    Service manual VX-400, ch. 7.3
  • Ask in your own words, on a tablet or phone.
  • Answer with its source: document, section, revision.
  • Where the data is unclear, the system says so instead of guessing.

Fewer calls to head office, more time at the machine.

Discuss this caseSee the use case

Every variant needs its manual, and from 2027 a digital one too.

Today

Variant VX-400-K12Documentation
  1. Operating manualcopied from VX-400open
  2. Engineering changes14 since rev. Bnot included
  3. Translations6 languagespending
  • Manuals are copied together again for each variant.
  • Translations lag behind the engineering state.
  • The Machinery Regulation adds new requirements.

With AI

Operating manual VX-400-K12Draft ch. 5

5.3 Coolant system maintenance

Maintenance only with the machine stopped and by trained staff.

Check filter F-3 every 500 operating hoursmaintenance plan

Check the coolant level daily at the sight glass.

Warning: high pressure, depressurise systemrisk assessment

Dispose of used coolant under waste code 12 01 09.

Spare part KP-12, item 7140-212parts list

  • Drafts come from BOMs, engineering data and existing manuals.
  • Variants stay consistent, changes carry through everywhere.
  • The writers check and release.

Documentation that keeps pace with engineering.

Discuss this caseSee the use case

The third identical fault only stands out if someone knows every report.

Today

Service reportsfree text
  • Report SB-26-0912MonSpindle leaking, seal replacedCustomer Hoffmann, VX-400
  • Report SB-26-0947WedLeak at spindle, retightenedCustomer Lenz, VX-400
  • Report SB-26-1003FriOil loss at spindle headCustomer Brandt, VX-450
  • Service reports and complaints exist as free text.
  • Only long-serving staff spot the patterns.
  • Engineering hears about it late, or not at all.

With AI

Cluster foundSpindle head seal
  1. Cases7 in 6 weekscluster
  2. SeriesVX-400, VX-4502 series
  3. Common factorSeal D-88, batch 2025-11suspect

Analysis to engineering

  • The AI finds similar fault patterns across all reports.
  • Clusters become visible, with evidence.
  • Engineering gets an analysis instead of a hunch.

Fix faults before they cost money for the tenth time.

Discuss this case

What it returns

Work it out for your service team.

Taking a service knowledge assistant as the example: how many people in service and on the hotline search how long for information? The calculation shows when the rollout pays for itself.

30
60 min
45 €

Assumption: AI halves the search time, at 18 working days a month. Costs at our prices: AI strategy and a €30,000 pilot once, operation at 20% of the project price a year. An example, not a promise. one-off setup €31,990, running cost per month €500.

Sources: Empolis, Service Express whitepaper: Weiss Umwelttechnik case; KVD, Insight-Report Service 2021; NBER Working Paper 31161, Brynjolfsson et al.; StepStone salary data 2026 and Destatis labour costs 2025

Hours freed per month
270
Saving per year
€145,800
Months to payback
3

Net after 24 months €247,610

Break-even
Check the calculation with your figures

We go through your values and tell you honestly whether it pays off for you. Your inputs are already in the request.

How we work

From AI strategy to everyday work.

Clarify first, then build, then run. Each phase has a fixed price, and after each one you decide again.

  1. Free intro call

    In 30 minutes we get a first impression of your business and work out whether an AI strategy is worth it.

  2. AI strategy

    Over one working day we go through your processes with you and your people, using real cases. The result is your prioritised roadmap: what comes first, what can wait, and what you are better off leaving alone.

  3. Pilot & implementation

    The first use case runs in your real operations within weeks, at a fixed price and with acceptance criteria agreed in writing up front. Built directly inside your systems.

  4. Operation & improvement

    We take on monitoring, model upkeep and keeping up with new rules such as the EU AI Act, and build the next case on the same base.

What it costs

Fixed prices, clear steps.

You know up front what each step costs and what it delivers.

  • Recommended start

    First call

    Free

    30 minutes by video

    • We look at where you stand
    • You find out whether an AI strategy is worth it
    • No obligation, with a concrete next step
    Free intro call

  • AI strategy

    €1,990plus VAT

    Fixed price

    • One workshop day on real cases
    • A calculation with your figures for the best use cases
    • A paper for management and shareholders

    Money-back guarantee: three use cases that pay back within 12 months on paper, with your figures, or your money back.

    Request the AI strategy

  • Pilot with one use case

    from €5,000plus VAT

    Fixed price

    • On your data, in your process
    • Acceptance criteria agreed in writing
    Request a pilot

All prices net. Operation afterwards is from 20% of the project price a year, cancellable monthly. The code and the data are yours.

Use cases for your industry

Where we start in your industry.

Use cases we selected for this industry, each in the systems you already use.

  • Service & engineering

    AI service assistant for technicians

    The AI service assistant answers questions from technicians and the hotline out of manuals, service reports and spare-parts lists, each with its source.

    • Ticketing
    • Manuals
    • ERP

  • Service & engineering

    Find spare parts by photo and serial number

    AI matches a request to the right spare part by photo, serial number or bill of materials and checks availability. The back office approves the quote.

    • ERP
    • BOMs
    • Parts catalogue

  • Sales

    Check specifications, prepare quotes

    AI reads specifications and tenders, matches requirements against your products and past projects, and prepares the quote for sign-off.

    • PDF
    • ERP
    • CRM

  • Customer service

    Sort fault reports and the service inbox

    AI captures fault pattern and machine from emails and forms, spots downtime and opens the ticket. Dispatch confirms the priority.

    • Email
    • Ticketing
    • ERP

  • Documentation & quality

    Create and translate technical documentation

    Technical documentation with AI: operating manuals and maintenance documents from engineering data and existing documents. Your editors release them.

    • PLM
    • DMS
    • Office

  • Sales

    Pre-fill customer questionnaires

    Pre-fills supplier, audit and sustainability questionnaires from your earlier answers and evidence.

    • Excel
    • DMS
    • Email

  • Documentation & quality

    Spotting repeat failures

    Finds patterns in service reports and complaints before the same fault reaches a customer for the third time.

    • Service reports
    • ERP
    • CAQ

  • Service & engineering

    Predictive service from your service data

    Predictive maintenance for machine builders who sell service: AI reads your service reports and shows which machines in your installed base carry the highest failure risk. No new hardware.

    • Service reports
    • ERP
    • Ticketing

Compliance from day one

AI that does not become a compliance problem.

Processing in German data centres, a data processing agreement from day one, and an EU AI Act risk class for every use case, documented before anything goes live.

Free intro callDiscuss your project

  • Germany

    Hosting and processing

    Run in German data centres under a data processing agreement, or directly inside your own infrastructure.

  • GDPR

    DPA under Art. 28

    The data processing agreement, the deletion policy and the technical and organisational measures are in place before the first line of code.

  • EU AI Act

    Transparent and documented

    We assign every AI system its risk class and document its purpose, data sources and limits in an auditable form.

  • Sign-off

    Human in the loop

    An agent proposes, your team decides. Anything binding that leaves the company, a quote for instance, is checked by a person first.

Further reading

All posts
MCP explained: how AI assistants reach your systems and what that means for your ITAI toolsTiemo Timtschenko
Feeding ChatGPT your own data? What a knowledge system does differentlyKnowledge systemsTiemo Timtschenko
Why the AI cannot find your part number: pgvector, Azure AI Search and others compared in plain termsKnowledge systemsTiemo Timtschenko

Frequent questions

What we are often asked.

The questions that come up in almost every first call, with our honest answers.

Where should a machinery builder start with AI?

Where most time is lost today searching and retyping. Usually that is quoting from specifications or service. The AI strategy shows which case pays off first for you.

Does this work with our ERP and PLM?

As a rule, yes. We connect the systems through their interfaces and keep the permissions that apply there. Anything that cannot be connected is settled in the AI strategy, before costs arise.

Does our engineering data leave the company?

Only if you want it to. We run the solution in German data centres or in your own environment and document every data flow up front.

What does the 2027 Machinery Regulation mean for our documentation?

The new EU Machinery Regulation applies from 20 January 2027. Among other things it governs providing instructions digitally. AI helps prepare and maintain the existing documents for that. It does not replace the legal review.

How long does a first AI project for a machinery builder take?

A first use case is usually in use after a few weeks. How long exactly depends on the data and the interfaces and is clear after the AI strategy.

How do you prevent wrong answers from the AI?

The AI answers only from your own documents and shows the source for every answer. If it finds none, it says so instead of guessing. Before go-live we test with real questions from your team and measure how often the answer is right.

Which manufacturing systems do you connect?

Typical are ERP systems such as SAP, proALPHA or abas, PLM systems, your DMS, SharePoint and Microsoft 365. Anything with an interface or an export can be connected. What cannot, we settle in the AI strategy.

Can everyone see every drawing and quote through the AI?

No. The AI takes over the permissions from your systems. Anyone who may not open a drawing or a quote today will not see it through the AI either.

Other industries

The next step

Let us find out where AI pays off for you.

30 minutes, free and without obligation.

Free intro call