It depends on the work, the systems it happens in and how sensitive the data is. Answer a few questions and you'll get an estimate you can check, not a figure from a pricing model.
Built for clients including Tata Steel, M&G Investments, Digistore24, Toppers in Concert, Bar-le-Duc and Loogman
Cost check
Work out what your AI solution costs
A few questions, then a request for an estimate you can check. No obligation.
Step 1 of 8Takes about two minutes
What drives the price
What drives the cost of an AI solution
An AI solution can be anything from an assistant that answers questions to a process that handles invoices on its own. These are the six things that determine the price.
What the solution needs to do
Answering a question is different from entering data in your accounts or changing an order on its own. The more the solution does itself rather than suggests, the more there is to build and test.
Volume and exceptions
Work that comes up many times a day needs a sturdier set-up than work that comes by now and then. Most of the effort is in the exceptions: unusual documents, incomplete emails, customers asking something other than expected.
Integrations with your systems
A solution that reads and writes in your email, CRM or accounts is only as good as that integration. The number of systems and the quality of their APIs account for a large part of the work.
The state of your data
An assistant on your own documents is only as good as those documents. If they are scattered, duplicated or out of date, tidying them up is part of the job. We'd rather see that up front than halfway through.
Privacy, approval and logging
Personal data calls for different choices than public information: which model, where it runs and whether hosting within the EU is needed. Human approval where it matters and logging of every action are always part of it.
Running and maintenance
After go-live you pay for language model usage, hosting and adjustments as the work changes. We work out those costs in advance, alongside what the manual work costs now.
From question to quote
How we arrive at a price
1
Your answers
The cost check tells us what you want built, who it is for and which parts it needs.
2
A short call
We call you back, ask what it needs to achieve and what is already in place, and tell you if an off-the-shelf package would suit you better.
3
A quote you can check
Scope, timeline and price on paper before anything is built. If the budget doesn't fit, we show you what to build first.
What an AI solution costs differs by department. Below, for each type of work, what drives the price, plus the questions we hear most often before someone starts.
1
How much does an AI solution cost?
The honest answer: it depends on what the solution needs to do. An assistant that lets staff search your own manuals is a different project from a process that reads invoices, checks them and enters them in your accounts. And much of what arrives as an AI question is really an automation question: if the rules are fixed, an integration is quicker and cheaper. We'll tell you that.
That is why we don't quote a figure on this page. With your answers from the cost check we give you an estimate you can check. In the quote we fix scope, timeline and price before anything is built, and set it against what the current manual work costs, so you see both sides.
2
How much does an AI solution for customer service cost?
The difference is in how far the solution goes. Sorting email and routing it to the right colleague is the smallest step. Preparing a draft reply based on your own information takes more. Looking up the order, changing it and answering the customer, with approval where it matters, takes the most building and also takes over the most work.
What else drives the price: how many channels questions come in through, how many systems are needed to answer a question, and whether customers talk to the AI directly or only your staff do. A model that answers customers directly is kept within tight limits: it answers only from your information and hands over to a colleague when it doesn't know.
3
How much does an AI solution for admin and invoicing cost?
Reading invoices, delivery notes and forms and entering the data in your accounting system or ERP lends itself well to AI, because the documents differ by sender. The cost depends on how many types of document come in, how easily your accounting package can be connected and which checks are needed before anything is posted.
If invoices already arrive in a structured format, such as e-invoices, you don't need AI but an integration. That is cheaper, and it's what we'll recommend.
4
How much does an AI solution for sales and quotes cost?
Here the work is in assessing requests, gathering the right information and preparing a quote. The price depends on how standardised your quotes are, where prices and terms come from and whether the solution needs to write to your CRM.
A quote goes past a person before it is sent: the AI prepares, your salesperson decides. We build that approval step in, and it is part of the price.
5
How much does an AI solution for knowledge and documents cost?
An assistant that answers staff questions based on your own manuals, procedures and case files. The model may only answer from those documents and shows the source with every answer. If it isn't in there, it says so.
The price depends mostly on the documents: how many sources, in which systems, how up to date they are and who is allowed to see what. If the permissions from your document management system need to carry over, so not everyone sees everything, that is a large part of the work.
6
Why quotes for AI vary so much
A demo on a handful of sample documents is quick to make. A solution that works on your own data every day, connected to your systems, with logging and approval where it matters, is something else. The difference between quotes is usually in what comes after the demo.
Ask whether the solution is tested on your own data, including the exceptions
Ask what language model usage costs per month
Ask where your data is stored and whether hosting within the EU is possible
Ask who owns the instructions, the integrations and the code
7
What does using a language model cost?
On top of the build you pay for usage. A model via an API charges per amount of text it reads and writes. An open model running within the EU mainly costs the server it runs on. How often the work comes up, how long the documents are and how heavy the model is together determine the bill.
Not every task needs the heaviest model. Sorting email can often be done with a small model; assessing a tricky contract needs more. Because we build model-independently, we choose per task, and you can switch later if prices or quality change.
8
Start small without throwing work away
Start with one process for one team, where the work comes up often. At first everything goes past a member of staff; what proves to work well may then run on its own. That way you see what it delivers before you invest further.
What we put in place in that first step stays: access to the systems, data in order, integrations, logging and an owner for each step. The second process builds on that and goes faster than the first.
9
What does an AI solution cost after go-live?
An AI solution isn't finished at handover. The work changes, documents are updated, systems get updates and questions come up that weren't in the tests. That's why we monitor, adjust and extend. Hosting and language model usage come on top.
We handle that maintenance as part of AI implementation, for a monthly fee that depends on scope and pace. We agree in advance what it covers, so you know where you stand.
10
How to keep costs under control
The most expensive AI project is one without a business case. So start with what the current work costs and where it gets stuck, and only then choose the technology. Sometimes the answer is an integration, sometimes an AI solution, sometimes nothing.
First work out what the manual work costs now
Start with the process that comes up most often
Leave fixed steps to regular automation
For each task, choose the smallest model that does it well
Amsterdam
Who you'll talk to
You talk to the founders
No salespeople. We read your answers and call you back. Then you build with a permanent team from our office in Amsterdam.
It depends on what the solution needs to do, how often the work comes up, which systems it happens in and how sensitive the data is. The cost check on this page lets you share that in a few minutes, and you get an estimate you can check. In the quote we fix scope, timeline and price before anything is built.
Why isn't there a price on this page?
Because a price from a pricing model is almost always wrong for AI. What a solution costs depends on your own data, systems and exceptions. We would rather give you an estimate based on your answers, from people who have read them, than a figure that sends you in the wrong direction.
What does AI implementation cost for a business?
With AI implementation you pay a monthly fee that depends on scope and pace. First we lay the foundations: access to the systems, data in order, integrations and logging. Then we automate the first process and build on from there. What exactly is included comes out of the AI Opportunity Audit, which has a fixed price.
Do I also pay for language model usage?
Yes. A model via an API is paid per use; an open model within the EU mainly costs hosting. How much that is depends on how often the work comes up and how much text is read per task. We work it out in advance and choose the right model for each task.
What does it cost to maintain an AI solution?
After go-live we monitor, adjust when the work or a system changes and extend where that pays off. Hosting and language model usage come on top. What that comes to each month depends on the scope, and we agree it in advance.
How long does it take to introduce an AI solution?
That depends on the number of systems, the state of your data and the number of exceptions. We start with one process for one team, so you see early on what it does. The timeline is in the quote.
Who owns the AI solution once it's finished?
You do. The code, the instructions, the integrations and your data live in your environment, with documentation. We build model-independently, so you aren't tied to one language model provider.
Is our data safe with an AI solution?
For each solution we agree which data the model may use, who can see what and where it runs. For personal data or confidential documents, the model can run within the EU. Every action is logged. As an organisation, Your Tech Club is ISO 27001 certified by Kiwa.
Cookies & privacy. We use functional cookies (always) and analytics and marketing cookies (only with your consent) to improve the site. Read our privacy policy.
