Equipment makers and industrial suppliers

AI worker teams for after-sales technical support

After-sales cost is decided by the accuracy of the first conversation. A wasted field visit dwarfs the cost of the call that caused it — which makes precision at intake the whole game.

EQUIPMENT MAKERS AND INDUSTRIAL SUPPLIERSCONSTRAINT ENVELOPE4 sector constraints designed to firstHUMAN AUTHORITY — NEVER AUTOMATED· Warranty and goodwill decisio…· Dispatch commitment· Safety judgement and design e…ACTS WITH APPROVALThresholds and gates on anything with cost or risk attachedACTS WITHIN POLICY5 blueprints start hereAutonomy widens outward only on evidence from the recordearned autonomy
The authority boundary we design to in Manufacturing — autonomy widens outward from the centre, and only on evidence.

Our view

Where we think the real opportunity is

Industrial after-sales has an information problem disguised as a service problem. The person describing the fault is usually not an expert, the person taking the call cannot read four hundred pages of revision-specific documentation in real time, and the gap between those two facts is filled with guesswork that becomes a dispatch. The leverage is therefore in grounding rather than conversation: identify the asset precisely, retrieve documentation matched to that exact model and revision, and refuse to guide when the match is not exact. That last part is the design decision that matters most. A near-matching manual is not a helpful approximation in industrial equipment — it is a wrong answer with safety and warranty consequences, so a near match is treated as no match and escalates. This also means the honest prerequisite is documentation quality, and that is worth confronting before anything is built.

Operating reality

How demand actually behaves here

Volume has a shape, and the shape is what breaks teams. This is what we assume about your operation before we design anything.

  • Faults are described by non-experts

    The operator reporting a fault often lacks the vocabulary to characterise it, so the useful diagnostic information has to be drawn out with structured questions.

  • Knowledge is fragmented across revisions and history

    Manuals, service bulletins, revision differences, and past case notes hold the answer between them, and nobody can search all of it under time pressure.

  • Distributors sit between you and the customer

    Support requests often arrive via intermediaries, which adds a translation layer and loses detail at every hop.

  • Demand is global and around the clock

    Equipment fails on the customer's shift, not yours, and downtime cost accrues while a support team sleeps in another time zone.

Where the hours go

The cost we would try to move first

Almost every operation spends its most expensive time on its least valuable work. Naming that precisely is what makes a pilot measurable rather than impressive.

  • Field visits that resolve nothing

    An engineer dispatched without the right diagnosis or part produces travel and labour cost plus a return visit, and it is the single most expensive after-sales outcome.

  • Wrong parts shipped

    Parts identified from an imprecise description generate freight both ways, extended downtime, and a second dispatch.

  • Engineers reading documentation

    Highly paid technical staff searching manuals for information that could be retrieved instantly is a persistent hidden cost.

  • Entitlement checked too late

    Warranty position established after expectations have been set creates commercial disputes that were entirely avoidable.

The trust envelope

What constrains the design, and what never gets automated

We design to the boundary first. A control added after launch is a control nobody believes — and in this sector the boundary is not negotiable.

Sector constraints

  • Revision-matched documentation only

    Guidance must come from documentation for the exact model and revision. A near match escalates rather than approximates.

  • Hard safety cutoffs

    Safety-critical symptoms end conversational troubleshooting on the turn they are recognised, regardless of how routine the rest of the call has been.

  • Entitlement verification before commitment

    Warranty and contract position is established before any timeline, cost, or dispatch is discussed.

  • Export and regulatory limits on parts

    Some components carry export-control or certification restrictions, and availability is not the same as permissibility.

Stays human, permanently

These are structural boundaries, not trust levels waiting to be relaxed.

  • Warranty and goodwill decisions

    Whether a repair is covered, and whether to cover it anyway, is a commercial judgement made by authorised staff.

  • Dispatch commitment

    The worker recommends; a person commits the engineer. Dispatch has cost and scheduling consequences that need accountability.

  • Safety judgement and design escalation

    Anything touching operator safety or a suspected design fault goes to qualified engineering immediately.

Where to start

Sequencing matters more than scope

The first blueprint should be the one whose success criteria your team already agrees on. Each step below links to the blueprint that implements it.

  1. Start

    One product family, asset ID and triage only

    Identify the asset by serial and revision, capture the symptom precisely, and route. No guidance is offered yet — this stage exists to prove the asset match is reliable, because everything else depends on it.

    Read the blueprint
  2. Then

    Approved guided diagnosis for that family

    Enable the approved procedure set, with safety cutoffs and entitlement checks live from the first day rather than added later.

    Read the blueprint
  3. Next

    Parts, dispatch briefs, and service booking

    Add parts checks and dispatch recommendations, measure first-visit fix rate against your baseline, then extend into planned-service scheduling.

    Read the blueprint

Systems we would expect to read and write

  • Asset and serial registry
  • Technical documentation library with revision control
  • Parts inventory and pricing
  • Field service management
  • Distributor and partner portals

Blueprints that apply

5 operating blueprints for Manufacturing

Each one carries the full team, process, and control detail — the mechanics this page deliberately does not repeat.

Push back on this

Objections we actually hear in Manufacturing

Answered as we would answer them in the room, including the cases where the right answer is that this is not for you.

Our documentation is not clean enough for this.
That is the most common blocker in this sector and we would rather surface it in week one than month four. The workable path is to scope one product family whose documentation is in reasonable order, prove the pattern there, and let the measured value fund the documentation work for the rest. If no family has usable revision-controlled documentation, the documentation project is the prerequisite and we would say so rather than build on sand.
A wrong troubleshooting step could injure someone.
Which is why safety cutoffs are enabled before any diagnostic capability, and why a near-matching manual is treated as no match. The worker's permitted actions are limited to approved procedures for the verified revision, and safety-critical symptoms end the conversation and route to a qualified person. If your safety review cannot get comfortable with that boundary, triage-only is a genuinely useful deployment on its own.
Our customers reach us through distributors, not directly.
That is often an argument for this rather than against it. The distributor layer is where detail is lost, and giving distributors a grounded intake path produces better-structured requests reaching you. It does need deciding up front whose knowledge and entitlement data the worker reads, which is a commercial question as much as a technical one.

Other sectors

How the argument changes elsewhere

Next step

Argue with this in a working session.

Bring your own numbers. We will map your version of the operating reality, agree what must stay human, and scope the first blueprint against a measurable win — or tell you it is not worth doing yet.