Industries7 min read
Manufacturer customer service: AI support in the aftermarket
AI can relieve a manufacturer's aftermarket by answering questions about manuals, spare parts, deliveries and warranty claims from reviewed knowledge and data from the ERP system. Technical assessments, warranty decisions and safety questions should go to a person. A good first step is to gather manuals and service notes in a knowledge base the team reviews.
Manufacturers get questions that are rarely difficult but often time-consuming: where the order is, which spare part fits, what the manual says and how a warranty claim works. The answers already exist, but scattered across the ERP system, PDF manuals and a few experienced people.
Why does the aftermarket get stuck on the same questions?
Because the knowledge sits in documents and with people, not where the customer asks. A dealer who wonders about a spare part has to send an email, and someone looks up the part number in a manual or in the ERP system. The same question comes back the following week from another customer.
The problem grows when the products have many variants and a long life. A machine sold today must be serviceable and have parts in ten years, so the answer has to apply to the right model and the right year. A general AI chatbot that guesses from what it has learned in general is the wrong tool here. The answer must come from the company's own reviewed material.
Which questions does a manufacturer get most, and who should answer?
The table shows typical examples and where they can be handled. It is not statistics from any company, but a way to sort your own questions.
| Typical question | Channel and knowledge | Needs an extra check |
|---|---|---|
| Where is my order, and when will it be delivered? | AI answer with data from the ERP system | Delivery deviations |
| Which spare part fits my model? | Help centre with reviewed parts lists | Older or custom-built variants |
| Where can I find the manual or assembly instructions? | Help centre and chat with a link to the right document | None, if the document is current |
| How do I make a warranty claim? | AI form that collects order number, photo and description | The assessment of the defect |
| Can you change the delivery address? | Action in Business Central, if you switch it on | Orders already shipped |
| Why have I received this invoice? | AI answer with invoice data from the ERP system | Disputes and credits |
| Is the product safe to use after the damage? | Always a person | The whole question |
What applies to warranty claims between businesses?
Between businesses, the Swedish Sale of Goods Act applies unless otherwise agreed. Under section 32 of the Sale of Goods Act, the buyer must give notice within a reasonable time after noticing the defect, and the right to rely on the defect is lost after two years unless a warranty or similar undertaking gives longer. If the products are sold to consumers, the Swedish Consumer Sales Act applies instead, with a three-year period for complaints.
This fits in a reviewed help article together with your own warranty terms. The assessment of whether a product is defective, and what should be replaced, is a decision a person makes. The AI's job is to collect the right information from the start, so the case does not have to go back and forth.
How does Supportifier solve it?
Supportifier gathers manuals, service notes and common answers in a knowledge base the team reviews. Up to 50 documents at a time, such as PDF manuals and Word files, can be imported and split into one article per topic. Contradictions between sources are flagged and everything goes to review before it is used.
The same reviewed knowledge is then used in the help centre, the chat, the AI form and the inbox, where the AI agent puts forward suggested replies that the team reviews and sends. If the material is not enough, the AI agent says so and hands over to a person instead of guessing.
With a connection to the ERP system, the AI agent fetches the customer's orders, invoices and deliveries. Supportifier has ready-made connections to Business Central and Visma Business NXT, listed on the integrations page. In Business Central the AI can also change a delivery address, remove an order that has not shipped, void an unpaid invoice or create a credit note, but only the actions you switch on, and you choose whether the AI suggests them or carries them out itself.
The help centre, chat and forms are available in Swedish, English, Norwegian, Danish, Finnish and German, and the AI answers in the customer's own language. That matters for a manufacturer with dealers in several countries.
What does Supportifier cost for a manufacturer? A worked example
Supportifier is priced by tickets per month, not per user. An example: a manufacturer gets 300 tickets a month from customers and dealers, and wants to connect the ERP system.
| Example | Per month | Per year with yearly billing |
|---|---|---|
| Start, 300 tickets per month | USD 89 | USD 959 |
| Connection to an ERP system with ongoing responsibility | Quote | Quote |
Knowledge base, help centre, chat, forms and inbox fit in Start for USD 89 a month. If Supportifier is to take ongoing responsibility for a connection to Business Central, Visma Business NXT or another ERP system, it is priced as Enterprise/ERP, with a written proposal before the start. Many therefore start with the knowledge and the inbox and add the connection once the first flow works.
Prices exclude VAT and are taken from the pricing page in November 2026. Every channel, every AI agent and unlimited users are included in each plan. Above the plan's volume it costs USD 40 per 100 tickets started, and you can change plan at any time. The volume in the example is chosen to show the calculation, not statistics from any company; calculate with your own figures.
Common questions
Can AI answer technical questions about our products?
Yes, if the answer is in a reviewed article, such as a manual or a service note. The AI agent answers only from material you have approved and shows which article the answer is based on. Questions that need a technical assessment, such as troubleshooting beyond the manual or safety after damage, should go to a person.
Can AI decide on a warranty claim?
No. The AI form can collect order number, photos and a description of the defect and explain your terms and the rules of the Sale of Goods Act. The assessment of whether the product is defective and what should be replaced is made by a person. That shortens handling because the information is complete from the start.
Does it work with our ERP system?
Supportifier has ready-made connections to Business Central and Visma Business NXT. Other ERP systems have no ready-made connection and are scoped during onboarding, via API, webhook or Retool. Ongoing responsibility for a connection to an ERP system is quoted as Enterprise/ERP.
Do we need to replace our current customer service system?
No, you do not need to start from that. An existing support system can be a source and a channel. Many start with the help centre and the knowledge base alongside what they have, and decide on the rest once the first flow works.
Next step
Want to try it yourselves first? Start 7 days free, no card needed. Would you rather walk through a first flow with us: bring a recurring question, where the answer lives today and which systems the team uses. Book a walkthrough.
Read more
- Customer service for energy, water and city networks with AI
- Automated returns, refunds and credit notes from support
- Order status from PostNord, DHL and Ingrid in every channel
Sources
Rules and product details checked on 21 November 2026. This page is general information, not legal advice.
This page is also available in Swedish: Kundtjänst för tillverkare: AI-stöd i eftermarknaden