SERVICESAI CUSTOMER SERVICEAI for your customer service, with a person where it matters
AI customer service that sorts questions, answers standard questions from your own sources and hands over the rest with context. In the mailbox, the chat, WhatsApp and on the phone, connected to your helpdesk and online store.
ISO 9001 & 27001
The agent does the work in your own systems. A person decides where it matters. Meet the AI employees →
What is AI customer service?
AI customer service is the use of AI to sort customer questions, answer standard questions and take over the preparatory work of a support team, in channels such as email, chat, WhatsApp and phone. Unlike a standalone AI chatbot, it works in the systems that are already there, such as the helpdesk, the online store or the ERP: it looks up an order, puts a ticket with the right team and hands a conversation over to an employee with the context attached. Your Tech Club builds AI for customer service that answers from an organisation's own sources, logs every action and leaves exceptions to a person. Your Tech Club is ISO 27001 and ISO 9001 certified.
Built for clients including Tata Steel, M&G Investments, Digistore24, Toppers in Concert, Bar-le-Duc and Loogman
Three situations we hear about from organisations that want to automate their customer service.
The same questions every day
Where's my order, how do I send something back, when are you open. The answer is already written down somewhere, but someone has to look it up and type it out again every time.
Reading, sorting and looking up first
Before anyone can help, the message has to be read, categorised and passed to the right colleague. That work takes time and gives the customer nothing yet.
The chatbot doesn't know and doesn't pass it on
A chatbot that only knows fixed answers leaves the customer stuck as soon as the question is different. That customer emails or phones anyway, and is already annoyed.
Recognise any of these?
An AI chatbot or AI across your customer service?
An AI chatbot for customer service is one channel. Customers also email, message and phone, and most of the work isn't in the answer itself but in what comes before it: reading, categorising, looking up the customer and their order. That's why we first look at all the questions coming in, not just the chat.
It often turns out that sorting and draft replies in the mailbox or the helpdesk deliver the most, and that a chatbot only makes sense after that. Sometimes a better FAQ or an integration with the online store is enough. Then we say so.
- First the mix of questions across all channels
- Sorting and draft replies often deliver the most
- A chatbot if it's still needed after that
Answers from your own sources, with a person on the exceptions
The AI only answers from what you have written down: your knowledge base, your terms and conditions, the data in your online store or ERP. Every answer shows its source. If something isn't there, it says so and hands over, rather than guessing.
For each type of question we set out what the AI may send by itself, what goes to an employee as a draft first and what always belongs with a person, such as goodwill gestures, complaints and angry customers. At first everything goes past your team. What demonstrably goes well then gets more room. Every action is logged.
- Answers only from your own sources
- What the AI may do by itself, set per type of question
- Goodwill, complaints and angry customers go to a person
Customer service automation in email, chat, WhatsApp and on the phone
Customers choose how they get in touch. That's why the AI works behind the channels you already have: the shared mailbox, the helpdesk, the chat on your site and WhatsApp through the WhatsApp Business Platform. For the phone we build an AI phone assistant that answers when it's busy or out of hours and books a callback or an appointment.
The overview and history stay in your own helpdesk, such as Zendesk or Freshdesk, and order status and returns come from your online store or ERP. We can connect any system. If you serve customers in several languages, the AI answers in their language from the same knowledge base as your Dutch team.
- Mailbox, helpdesk, chat and WhatsApp
- A phone assistant for busy times and evenings
- Answers in the customer's language, from one source
AI that works in your own systems, with the control that customer contact calls for.
We start with your mix of questions
Which questions come up most, what do they cost and where is the answer already? That decides where AI starts, and whether a chatbot is needed.
In the systems you already have
Your helpdesk, online store and ERP remain the source. The overview stays in one place, and nobody has to retype anything.
A person where it matters
What the AI may do by itself is set per type of question. Complaints, goodwill and doubtful cases go to your team with context.
Every answer with a source, every action logged
What the AI saw, answered and did can always be traced afterwards, even when nothing went wrong.
Yours, not a vendor's
The instructions, the integrations and the knowledge base remain yours. We are model-agnostic, so you aren't tied to a single provider.
We stay until it works
After go-live we keep an eye on what the AI passes on and why, keep the knowledge base up to date and expand to the next channel.
- projects delivered
- returning clients
- from 127 reviews
- ISO27001 and 9001 certified
Three ways to bring AI into your organisation, side by side.
- Works in your own systems, such as ERP, CRM and emailStandalone AI tool: partly · AI freelancer: yes
- ISO 9001 and ISO 27001 certifiedStandalone AI tool: varies · AI freelancer: rarely
- Starts with the process, not the technologyStandalone AI tool: no · AI freelancer: varies
- Tells you when AI isn't the answerStandalone AI tool: no · AI freelancer: varies
- A person decides where needed, with loggingStandalone AI tool: partly · AI freelancer: varies
- Strategy, design, software and AI in one teamStandalone AI tool: no · AI freelancer: no
- Management and further development after go-liveStandalone AI tool: via subscription · AI freelancer: varies
- Dedicated contacts, office in AmsterdamStandalone AI tool: no · AI freelancer: varies
Five steps. How long it takes depends on the number of channels and systems and on the state of your knowledge base.
- 01
Map the mix of questions
We look at a period of real questions: which come up most, how long they take and where the answer already is.
- 02
Set the sources and the limits
Which knowledge base, terms and systems the AI may use, what it may send by itself and what always goes to an employee.
- 03
Connect and test on real questions
We connect the AI to your helpdesk, online store or ERP and test it on questions from your own day-to-day work, including the difficult ones.
- 04
Run alongside with draft replies
At first the AI only prepares drafts that your team reviews. What demonstrably goes well may then run on its own.
- 05
Go live and expand
Live with monitoring on one channel. If it works, channels, languages or types of question are added on the same foundation.
After go-live
- MonitoringWe track what the AI handles by itself, what it passes on and why, and see it when something deviates.
- Keeping the knowledge base up to dateQuestions without a good answer become visible, so your team knows which article is missing or out of date.
- Access and loggingPermissions follow the roles in your team, and logging stays on, even when everything goes well.
- ExpandingA next channel, an extra language or the phone, as soon as the first channel runs well.
No account manager. Your first conversation is with someone who does the work, often with one of the founders. After that you work with a dedicated team from our office in Amsterdam.
- Office on the Duivendrechtsekade in Amsterdam
- ISO 9001 and ISO 27001, certified by Kiwa
- 500 projects delivered
- 4.8 from 127 reviews
AI agents that work in the ticketing, online store, CRM and ERP systems that were already there.
Connected to what was already running
For a retail organisation we built AI agents that work in the existing ERP, CRM, online store and ticketing systems, with an overview layer in the organisation's own Azure tenant.
Every action traceable
What the agents do goes into an append-only audit log. So it is always possible to check afterwards what happened and why.
What can AI handle by itself in customer service?
Questions whose answer is in your own sources, such as order status, opening hours, returns policy and instructions. Plus the preparatory work: sorting messages, assessing urgency, looking up the customer and their order and preparing a draft reply. Exceptions, goodwill gestures and angry customers go to an employee.
What is the difference between an AI chatbot and AI in customer service?
An AI chatbot for customer service is one channel: it answers questions in a chat window. AI in customer service works behind every channel, in your mailbox, your helpdesk, WhatsApp or on the phone. It looks up the order, puts the ticket with the right team and hands over with the context attached. A chatbot can be part of that.
What does AI for customer service cost?
That depends on the number of channels, the systems the AI has to work in and how much has to go past an employee. An assistant that sorts email and prepares draft replies is a different project from a helpdesk in four languages with the phone added. We build it as part of AI implementation, for a monthly fee that depends on scope and pace. On the cost page for AI agents you get a reasoned estimate for your situation.
How long does it take to put AI into customer service?
That depends on the number of channels and systems and on how well your knowledge base is kept. We start with one channel or one type of question, let the AI first write draft replies that an employee reviews, and only give it more room once that demonstrably goes well. The timeline is in the quote.
How do you stop AI from giving a customer a wrong answer?
By letting it answer only from your own sources and showing the answer with its source. If something isn't there, it says so and hands over to an employee. That's why keeping the knowledge base up to date is part of the setup: an outdated article gives an outdated answer.
Does AI customer service work with Zendesk, Freshdesk or Shopify?
Yes. The AI works in the system you already have, so the overview and history stay in one place. Order status and returns come from your online store or ERP. We can connect any system, including bespoke or older software.
Can AI also answer the phone and WhatsApp?
Yes. An AI phone assistant answers when it's busy or out of hours, handles standard questions and books an appointment or a callback. The caller hears at the start that they are speaking to an AI assistant. WhatsApp runs through the WhatsApp Business Platform, with retention periods agreed in advance.
Can AI help customers in other languages?
Yes, from the same knowledge base as your Dutch team. We agree in advance which languages you really serve and what happens if a customer keeps asking in a language nobody can take over in. The conversation then goes to an employee with a translated summary.
Where do you start with customer service automation?
With the questions coming in now: which come up most, how long they take to answer, and where the answer already is. If you don't know that yet, the AI Opportunity Audit works it out for you in five working days, and also tells you where AI doesn't pay off.
Related services
Get in touch with us directly. We'll get back to you as soon as we can
Your message goes straight to our team in Amsterdam.