Manufacturing · Customer service

Every fault report sorted before anyone opens the email.

Fault reports from the service inbox are sorted by the AI by machine, error pattern and urgency. Downtimes rise to the top, dispatch confirms.

Free intro callDiscuss this case
Service inbox12 new
  • Hoffmann Stanztechnik06:42Line down, E-217Not running since 6 amDowntime
  • Kunststoffwerk Lenz07:15Maintenance in MarchPlease suggest a dateAppointment
  • Metallbau Brandt07:58Noise at spindleStill running, getting louderFault
Ticket SV-26-0388Ticketing
MachineVX-400 · SN 40218
FaultE-217, coolant pressure
PriorityDowntime, immediate
Confirm

Where time goes today.

  1. Downtime gets lost in the inbox

    The urgent report sits between appointment requests and invoice questions. By the time someone reads it, the line has been down for an hour.

  2. Key details are missing

    Serial number, error code, photo: often exactly what the technician needs is missing. The follow-up costs half a day.

  3. Retyping eats into dispatch

    Every email is copied into the ticketing system by hand. That time is missing for planning visits.

One case, step by step.

Service inbox12 new
  • Hoffmann Stanztechnik06:42Line down, E-217Not running since 6 am
  • Kunststoffwerk Lenz07:15Maintenance in MarchPlease suggest a date
  • Metallbau Brandt07:58Noise at spindleStill running, getting louder

Every report lands in one place.

Service inbox, contact form and customer portal. The AI reads text, attachments and photos.

Your team decides

Which channels are connected.

Report recognisedHoffmann Stanztechnik
  1. MachineVX-400 · SN 40218found
  2. Error codeE-217, coolant pressurefound
  3. StatusLine down since 06:00downtime
  4. ContractService Premium, 4 hresponse time

Machine, fault and urgency recognised.

The AI matches the report to the machine in the customer file, reads error code and fault and recognises whether the line is down.

Your team decides

What counts as downtime and which contracts take priority.

Queueby priority
  • Hoffmann Stanztechnik06:42Line down, E-217Respond by 10:42Downtime
  • Metallbau Brandt07:58Noise at spindleQuestion: serial numberFault
  • Kunststoffwerk Lenz07:15Maintenance in MarchDate suggestion readyAppointment

The queue sorts itself.

Downtime with response times at the top, appointments and questions behind. Where details are missing, the follow-up question is ready.

Your team decides

The order in which work is done.

Ticket SV-26-0388Ticketing
CustomerHoffmann Stanztechnik
MachineVX-400 · SN 40218
FaultE-217, coolant pressure below 2.5 bar
PriorityDowntime, respond by 10:42
Confirm

The ticket is filled in.

Machine, fault, contract data and a suggestion for the visit are in the ticket. Dispatch confirms and plans.

Your team decides

Priority and who goes out.

What it delivers and what it connects to.

  • +14%

    more issues resolved per hour when customer service has AI support

    Source: NBER Working Paper 31161

  • < 5 min

    until downtime is at the top of the queue

    Assumption for a typical company

  • 2 h

    less retyping per dispatcher and day

    Assumption for a typical company

Works with the systems you already use

Who does what.

What the AI takes on

  • Reads emails, forms and attachments
  • Recognises machine, error code and downtime
  • Sorts by urgency and contract
  • Drafts questions for missing details
  • Creates the ticket as a draft

What your team decides

  • Confirms the priority
  • Plans the visit
  • Talks to the customer
  • Decides on special cases

Who it fits.

A good fit if

  • Fault reports arrive by email, form and portal in the service inbox
  • Machines and service contracts are on the customer record
  • Your contracts set response times
  • Dispatch types reports into the ticket system by hand today

Less of a fit if

  • Only a few reports come in each day
  • Faults are reported almost only by phone
  • Reports go straight to service partners, not to you

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.

Built fromWorkflow automationAI integration

Further reading

All posts
MCP explained: how AI assistants reach your systems and what that means for your ITAI toolsTiemo Timtschenko
Why the AI cannot find your part number: pgvector, Azure AI Search and others compared in plain termsKnowledge systemsTiemo Timtschenko

Frequent questions

Common questions about it.

The questions that almost always come up about this use case.

What if the AI misjudges downtime?

It always shows what it bases the urgency on. When in doubt it ranks higher, and dispatch corrects with one click.

Does the AI answer customers itself?

Only if you want it to, for example with an acknowledgement. Answers on substance go through your team.

Does this work with our ticketing system?

Common systems such as Zendesk, Freshdesk, SAP Service or Jira connect through their APIs. For in-house tools we check the route in the AI strategy.

More use cases

The next step

Let us find out where AI pays off for you.

30 minutes, free and without obligation.

Free intro call