Abuscom
← Back to client projects
AI customer support · ISA

Customer support with AI – together with an industrial manufacturer.

Good customer support is just as important as the product itself. Together with an industrial manufacturer we are building, step by step, a system that unlocks existing knowledge and makes it directly available to support.

Client
International industrial manufacturer
Industry
Industrial measurement and scanning technology
Services
AI consulting · Development · Integration
Timeframe
2025 – ongoing
Support environment at an industrial manufacturer
What it's about

Better support around complex machines.

Our client designs and builds highly complex industrial scanning and measurement systems that are in use with customers around the world. Machines like these run around the clock and are integrated directly into their customers' production. Downtime therefore often means a standstill.

That is why good customer support is just as important as the product itself. Together with the company we are working to noticeably improve the support around these machines – so that requests can be handled faster, more precisely and less dependently on individual people.

01 — The starting point

Support runs on experience – which is hard to grasp.

Behind the question of how a particular problem arises and, above all, how it is solved lies a great deal of knowledge and experience. That knowledge exists – spread across resolved tickets, technical documentation, experienced staff and further systems – and is therefore hard to grasp.

The company serves customers worldwide. Requests reach support in different languages and time zones. The lever therefore is not more staff, but making existing knowledge more readily available.

The challenge
  • Customers worldwide, in many languages
  • Knowledge spread across many systems
  • Barely retrievable during the live conversation
  • Strongly tied to individual people
02 — Our path

A path in several stages.

Instead of a finished, isolated solution, we are building a system together with the client step by step – one that delivers concrete value at every stage and builds on the previous one. At its core is ISA – a system that unlocks existing knowledge and makes it available where it is needed: right in the support conversation.

Stage 1
01

A shared knowledge base

ISA brings together existing data that used to sit apart: past support requests, technical documents and even source code. From this emerges a searchable knowledge base that reflects the collected experience. Screenshots and error images from earlier cases are analysed too, so their content becomes findable – not just the written text. ISA fits into the existing system landscape, such as the current ticketing system, rather than replacing it.

Stage 2
02

Knowledge in the conversation, in real time

On this basis, ISA helps exactly when it counts: during a live request. The conversation is transcribed live, the actual issue is recognised automatically, and ISA suggests matching solutions from comparable cases – while the conversation is still going on. Every suggestion points back to its source, so it stays clear what a recommendation is based on. The search for the right information falls away; staff can focus on solving the problem.

Stage 3
03

Across language barriers

Because the company serves customers in many languages, we are working on translating conversations in real time. The goal is for everyone in the conversation to be able to speak in their own native language and still be understood. Language is no longer a hurdle – and support becomes possible regardless of who happens to speak which language.

Outlook

AI agents at first contact

Longer term, it is conceivable to let AI agents take initial support requests before a staff member takes over the conversation. Simple matters could be resolved directly, while the team focuses on the demanding cases. This step builds seamlessly on everything created in the stages before it.

Data sovereignty from the start

Support data is sensitive. That is why the AI models in use run on local hardware at the client's premises. The data does not leave the premises and is not passed on to external services. This makes it possible to reap the benefits of AI without giving up control over your own data.

03 — What the client gains

Support that noticeably relieves the team and gets better.

Faster

Relevant knowledge from thousands of cases is available within seconds during the conversation, instead of being scattered across archives.

More precise

Requests are handled more precisely and at consistent quality – regardless of who is on the phone.

More independent

Knowledge is no longer tied to individual heads. New staff also get up to speed more quickly.

Without language barriers

Customers are supported in their own language – worldwide.

Data-sovereign

Sensitive support and customer data stays entirely in-house.

Do you face a similar challenge?

Let's talk, with no strings attached, about where AI delivers real value in your company.