AI in everyday workflows
AI integrations for business
AI integration is connecting an AI model to your website, documents or workflows so that it carries out a specific task. We build AI chatbots for websites, internal AI assistants and n8n automations in which a person checks the outcome before it is sent.
On this page 7 sections
What can you do with AI integration?
For a small business, the most useful AI tasks are usually simple: answering recurring questions, sorting enquiries, extracting data from documents and preparing a draft for a person to review. We start with one clear task whose result can be checked.
- 01
AI chatbot on your website
Answers customers using the information you provide: services, prices, delivery terms and frequently asked questions.
- 02
Internal AI assistant
Finds answers for your staff in your instructions, contracts or product descriptions.
- 03
Sorting enquiries and emails
Recognizes the type of enquiry, routes it to the person responsible and prepares a draft reply.
- 04
Extracting data from documents
Extracts fields from invoices or orders that can be transferred to Rivilė, B1 or another accounting system.
- 05
Content drafts
Drafts of product descriptions or reply templates, edited by a person before publishing.
AI chatbot or AI assistant: what is the difference?
An AI chatbot works publicly on your website and talks to customers. We give it only information that can be made public and set when the conversation is handed to a person, for example when a customer asks about a specific order or a complaint.
An AI assistant is for your staff. It can use internal documents, so access matters most here: who can use it, which documents it sees and where conversations are stored. In both cases we first agree which questions the solution answers and which it does not.
| AI chatbot | AI assistant | |
|---|---|---|
| Who uses it | Your customers, publicly on the website | Your staff |
| What information it uses | Only what can be made public | Internal documents, according to the access granted |
| What we agree first | When the conversation is handed to a person | Who can use it, which documents it sees and where conversations are stored |
| Price | By scope, €45/hour | By scope, €45/hour |
How we connect AI with n8n
n8n is a workflow automation platform where every step is shown as a diagram: where the data comes from, what the AI model does with it, who checks the result and where it is sent. We use it when AI has to work together with other systems.
n8n can run on your own server, or you can use the n8n cloud version. We choose between them based on what data the automation will process and who needs to see it.
- 01
Email and forms
A new enquiry from a website form or a mailbox starts the workflow.
- 02
CRM, Google Sheets and accounting
The result is saved where you already work, not in yet another separate tool.
- 03
WooCommerce and other online stores
Order, product and customer data is transferred through the API.
- 04
Human approval step
Before a reply is sent to a customer or data is saved, the person responsible reviews it.
What can AI integration not guarantee?
AI models make mistakes: they can misunderstand a question, miss a field in a document or give an answer that sounds convincing but is wrong. That is why we design checks, limit the topics the solution answers and log its actions. We do not promise a specific amount of time saved or sales growth. It can only be assessed after testing the solution with your data.
Data and privacy
Before we start, we find out what data will go to the AI model provider. If the task does not need personal data, we do not send it to the model. How the provider handles the data it receives depends on its terms and the chosen plan, so we review them together before the project.
If the solution will process personal data of customers or staff, your company remains the data controller. We recommend discussing the legal GDPR compliance assessment and contracts with providers with a lawyer.
How much does AI integration cost?
We have no fixed AI packages, because projects differ a lot: a simple chatbot with a few hundred questions and answers is very different work from an automation that reads invoices and passes the data to accounting. After assessing the task, we give you the scope of work and the price in writing. Work is charged at our €45/hour rate.
Third-party fees are paid separately: the usage fee of the AI model provider, such as OpenAI or Anthropic, and, if needed, an n8n or server fee. We list them in the proposal. If you are not yet sure where to start, a consultation at €45/hour is a good fit.
How an AI integration project works
The first stage is small and testable, so you can see how the solution works with your real data before a larger investment.
Task description
What work the AI has to do, who will check the result and what a good answer looks like.
Pilot stage
We test the solution with your examples and write down errors and edge cases.
Launch with checkpoints
The solution goes live with a human check, logs and a clear procedure for switching it off.
Monitoring and adjustment
We review the answers, add to the knowledge base and agree who will look after the solution from then on.
Frequently asked questions
How much does an AI chatbot for a website cost?
We charge for building an AI chatbot at the €45/hour rate and give you the final amount in writing after assessing the task. The price depends on how much information the chatbot needs to know, whether it connects to other systems and when it hands the conversation to a person. The AI model provider's usage fee is paid separately.
Can an AI assistant answer customers without a person?
Yes, for clear, recurring questions such as opening hours, delivery or the price list. For orders, complaints, contracts and other sensitive cases we recommend handing the conversation to a person. Before launch we agree which topics the AI handles and which it hands over, and record this in the settings. Customers can always ask to speak to a person.
Which AI models do you use?
We mostly work with OpenAI and Anthropic Claude models through their APIs, and connect the workflows in the n8n platform. We choose the model by task, language, cost and data requirements. For Lithuanian texts we always test the model with your examples, because the quality of answers differs between models.
Will our data be used to train AI models?
Usually not: the OpenAI and Anthropic business API terms state that data sent through the API is not used to train models by default. Before the project we review the terms of the chosen provider and plan together with you. We send the model only the data the task needs, and recommend discussing personal data questions with a lawyer.
Describe the task you want to hand over to AI, and we'll tell you whether it is a good fit.
We reply with a proposed first stage and a clear price. You can also write to us directly:
info@techprojects.ltDiscuss AI integration