Automate customer requests: AI sorts, the process sets the limits

swoox.io assigns every incoming request to a fixed category and forwards it to the responsible team. Where the AI agent is unsure, a person decides.

Industry: All

Department: Customer Service

Table of Contents

When request management starts in a shared inbox
How swoox.io automates the assignment of customer requests
What is the agent allowed to do?
Email management in customer service, step by step
AI in customer service: what changes day to day
Why swoox.io for SMEs?
FAQ: automating customer requests with swoox.io

When request management starts in a shared inbox

The working day in service starts in the shared inbox. Billing questions, appointment requests, complaints and delivery enquiries all come through the same channel. Before the first request is handled, time goes into sorting, forwarding and asking who is responsible.

How a request is assigned depends on who is reading. An urgent request with a harmless subject line ends up with the wrong team and waits there. If volume rises or someone is on holiday, the person who knows where everything belongs is missing. When someone asks later, it is hard to reconstruct why a request ended up where it did.

How swoox.io automates the assignment of customer requests

Every new request starts a process. An AI agent reads the free text and assigns category and urgency. It chooses only from a catalogue defined by the business team. The process then routes the request to the right target path. Customers receive an acknowledgement based on an approved template. If the assignment is uncertain, a person checks it before the request moves on.

Technically, this is a BPMN 2.0 process. The AI part sits in an agent task with a clearly defined assignment and a fixed output schema. The business team maintains categories and target teams in the low-code editor. The handover to existing systems such as a ticket system is set up in the project. You can connect ChatGPT, Claude or Llama as the model. If the model changes, the process stays the same.

What is the agent allowed to do?

Platform Level

The tool allowlist and permission system determine what the agent can access. It reads the request text and the category catalogue. It may not reply to customers, change data in other systems or delete requests.

Process Level

The BPMN process is an explicit graph. The agent chooses from fixed categories and cannot invent a new one. Where a category leads is defined in the process model and not left to the agent’s judgement.

Agent Level

If the confidence of an assignment is below a defined threshold, the agent stops and hands over to a person. At the consequence boundary, a person always decides – wherever a decision directly affects the customer.

Every assignment by the agent is recorded in the process log and stays traceable. This setup matches the approach that researchers at Fraunhofer FIT describe for the controlled use of AI agents in business processes.

Email management in customer service, step by step

(Example workflow – each process is unique)

A request arrives via form, interface or inbox and starts the process.
The AI agent reads the free text and assigns category and urgency from the fixed catalogue.
A person decides here if the assignment is uncertain. Otherwise, a wrongly sorted request would sit with the wrong team.
A gateway routes the request to the target path defined in the process model for this category.
The customer receives an acknowledgement based on an approved template.
The target team receives a structured handover, directly in the existing ticket system if required.
Every step is documented in the process log.

AI in customer service: what changes day to day

Every request is assigned using the same criteria, no matter who is on duty.
The team starts by handling requests instead of sorting them.
Urgency comes from the content of the request, not the subject line.
Uncertain cases go to a person instead of the wrong team.
Customers receive a consistent acknowledgement.
For every request, you can show when and why it went where.
The business team maintains new categories or target teams itself.

Why swoox.io for SMEs?

swoox.io runs the way that fits your IT: on-premise in your own data centre or as PaaS in the EU. Even a small IT team can manage SSO/MFA and the role and permission system on its own. Customer data stays where the company wants it.

Assigning incoming requests is often the starting point in customer service. Further processes follow on the same platform, such as ticket management or complaint handling. Engine and governance stay the same, and modelling is done with low-code based on BPMN 2.0.

There is a clear line for integrations. swoox.io includes REST interfaces, email sending and templates as standard building blocks. Connecting a specific ticket system or CRM is part of the implementation project.

FAQ: automating customer requests with swoox.io

Can customer requests be automated without the AI replying itself?

Yes. The agent assigns, it does not reply. Customers only receive an acknowledgement based on an approved template. The responsible team writes the actual reply.

Does swoox.io replace the ticket system for request management?

No. swoox.io sorts the requests and hands them over to your existing system in a controlled way. The integration is set up in the project.

Who decides on ambiguous requests, such as a complaint with a billing question?

If the agent is unsure, it stops and hands over to a specialist. They confirm or correct the category, and only then does the process continue.

Which models does swoox.io use for AI in customer service?

ChatGPT, Claude or Llama, depending on your requirements for data protection and operations. Changing the model changes neither the categories nor the process.

Do customer requests and customer data stay in the EU?

Yes. On-premise or PaaS in the EU – you decide based on your operating model. For a demo, sample requests are enough, so no real customer data is needed.

More demo cases of Customer Service

Ticket management with AI agents: deadlines in view, escalation to people

Department: Customer Service Industry: All

swoox.io takes over the work around your existing ticket system. An AI agent classifies every ticket using fixed categories. The process monitors SLA deadlines and escalates to the responsible person in time.

View Demo Case

Complaint management with AI agents: prepared by AI, decided by people

Department: Customer Service Industry: All

swoox.io puts every complaint into a fixed process. An AI agent classifies the case and suggests a solution. A person decides on goodwill, and every step stays traceable.

View Demo Case

Get to know swoox.io

We’d be happy to show you how easy and flexible process automation with swoox.io can be.
Or you can try it out for yourself.

Try it for free

Request a Demo