Blog · Published · Updated
How to choose an AI implementation partner
Short answer
Choose an AI agency that walks through one real case from your process with you, shows a hard case live and describes acceptance and maintenance before you sign. Compare the price of the same result. At Syntalith an audit costs from €600 net and we credit it towards the build. Automations, apps and agents start at €3,000 net.
- net for implementation for a team of up to 10 people, paid once
- from €1,150
- from start to work of the entire team
- 2 weeks
- after this time, we measure the effect on the same tasks
- 30 days
- conversation with an engineer and a written plan within 2 business days
- €0
We belong to the OpenAI and Anthropic partner programmes and learn through their official courses. What this means ↓
Author: Syntalith · 4 min read
Statistics
AI implementation in companies: statistics 2026
€600
net is the starting price of a Syntalith audit: a process map, the architecture and a fixed quote. If you order the build, we credit it towards a rollout that costs €3,000 to €17,500 net.
40%
According to Gartner (2025), over 40% of agentic AI projects will be cancelled by the end of 2027 because of costs, unclear value or weak risk controls.
130
According to Gartner (2025), of the thousands of agentic AI vendors only about 130 offer real agentic solutions.
70.9%
According to Eurostat (2025), 70.9% of EU companies that considered AI and did not use it named a lack of relevant expertise.
Source: Eurostat, Use of artificial intelligence in enterprises (2025) ↗
48%
According to the European Commission (2026), 48% of EU SMEs and mid-caps name a lack of expertise as a barrier to AI.
Source: European Commission, Flash Eurobarometer 578 (2026) ↗
Figures checked: 22 October 2026
Start with one real case
Prepare one case from daily work and walk through it with the candidate's engineer. Show where it comes from, where an employee looks for information and what delays the finish. A case with missing data or conflicting instructions works best.
Judge the questions. A good supplier asks about the result, access to systems and the person who takes the decision. Their answer may lead to an agent, a simple automation, putting knowledge in order or training. If every conversation ends with a proposal for an agent, the candidate is selling one product. Also ask who will work on the project. The person from sales is rarely responsible for integrations and later support.
The process fits ready-made connections and the team handles exceptions
a platform specialist, for example n8n
configuring the flow
You need a review panel, roles and custom logic
a company that builds software with AI
an app or an agent
The problem has no name yet
a process audit before choosing a supplier
from €600
Give every candidate the same brief. Later, carry the criteria from the brief over to the acceptance card.
Read the portfolio by similarity of problem
A similar industry helps, and it is not enough. Compare the type of data, the system's permissions and the way the user works. An agent that writes a draft reply and a system that changes an order are different purchases.
Tell work for a client apart from a test on synthetic data. Both show skill, and they answer different questions. When a client's material is confidential, the supplier can show a safe example or explain the assessment method. To check references, use the verification card.
Compare specific commitments
| Area | What to settle before you choose |
|---|---|
| Scope | the result of the process, integrations, work on the team's side |
| Responsibility | who designs, implements, assesses and maintains |
| Data | sources, access rights, place of processing, retention |
| Acceptance | test cases, scoring, what happens after a weak result |
| Maintenance | support hours, reporting failures, the scope of changes |
| Handover | code, documentation, accounts and data as the contract says |
A slogan about GDPR compliance does not replace agreements about suppliers and access. Do not use mechanical scoring either. One missing permission for an integration can weigh more than ten good answers.
Ask for a demo of a hard case
At the presentation, choose a situation where a document is missing, two pieces of information conflict or a system does not respond. Check what the user will see and how they pick up the work. Anthropic studied how users supervise the independent work of agents. For the buyer, one question follows: will the team see the course of the work and have time to react?
- The same case from your process for every candidate
- A missing document, conflicting data and an error in an external system
- A repeated message and a failed write to the CRM
- The history of actions and the point where a person decides
Compare the price of the same result
Ask for a quote on one case, for example: an email enquiry reaches the CRM as a draft quote with a list of gaps. Include access to the CRM, the price source, human review, tests and failure handling.
Process review
30 min, a written plan within 2 working days
€0
An engineer counts the cost of manual work and shows what to do first.
Test on your process
5 working days
€0
The system works on 20 to 100 cases the team has already handled.
Audit
5 working days
from €600
A process map, the architecture and a fixed quote. If you order the build, we credit it towards the rollout.
ChatGPT or Claude for the team
2 weeks
€1,150 to €4,600
Accounts, rules for working with data and training. Up to 10, up to 40 and above 40 people.
AI automation
2 to 6 weeks
from €3,000
A process with fixed steps: the inbox, documents, the CRM, reports.
AI app
4 to 10 weeks
€3,000 to €7,000
A tool for the team, for example document search that shows the source.
AI agent
6 to 16 weeks
€3,000 to €17,500
It does routine steps itself, hands exceptions to a person and records everything.
These are the prices of our services. Compare licences, the model, hosting and maintenance separately. According to Gartner (2025), over 40% of agentic AI projects will be cancelled by the end of 2027, most often because of costs and unclear value. According to Eurostat (2025), 70.9% of EU companies that considered AI and did not use it named a lack of relevant expertise. That is why you write the measure of success into the contract before the start.
How to assess a Syntalith proposal
Syntalith holds OpenAI Select Partner status in the OpenAI Partner Network and is a member of the Claude Partner Network. That tells you about the ecosystem, and a test on your process decides the choice. We build AI agents, automations, apps and training, so the form of the project follows from the problem. In the use case descriptions you can check the context of the systems and the scope of the evidence. For the free process review, bring a description of the process, the tools you use and one hard case.
„I already use AI.”
Your people too. Just no rules.
From our conversations with companies
5 na 10 employees paste sensitive data into private AI accounts: contracts, invoices, customer data.
In ESET survey (2026)
1 na 10 admits this in the survey.
Customer data pasted into a private AI account without an data processing agreement is a violation of the GDPR.
Don't worry, we won't tell anyone :)
What do we see in companies that "already use"
- Only ChatGPT or Claude accounts, without plugins, skills, integrations and MCP servers with company tools.
- Company programs are not connected to AI. People still click through systems and Excel, and enter chat data manually.
- Lack of company standards: everyone writes instructions in their own way, no one shares what works.
- No rules for personal data, GDPR and company secrets.
- No limits and no visibility: no one knows who uses AI, how much and for what.
62% Polish employees use AI at work · 27% companies have written rules for working with AI.
Source: ESET and DAGMA, Cyberportrait of Polish business 2026 (March 2026, 1,026 people) ↗
ChatGPT or Claude space review
Accounts, permissions, rules for working with data and 5 tasks set for the team. 1 day.
€450 net
Company accounts, rules of working with data and team training in 2 weeks, from €1,150 net.View implementation for the team →
„Is my data safe?”
Yes. You choose where they go: from your company account to the model on your server.
00Employee's private account
Customer and company data ends up in an account that the company cannot see or control. There is no data processing agreement, so it is a violation of the GDPR. When an employee leaves, the chat history goes with them.
01ChatGPT or Claude corporate account
The provider does not use your data to train models. A copy of the queries remains in his logs for up to 30 days.
02API with zero retention (ZDR)
The provider processes the request and does not save anything. It requires his consent, we prepare the application together.
03Processing in the EU
Queries are counted and stored in Europe: OpenAI in the EU region, Claude via clouds with the EU region.
04Model at your place
The open model runs on your server or in your cloud in the EU. The data does not go outside the company.
What we do with every implementation
- Confidentiality agreement and data processing agreement before we receive any data.
- You choose the cases for the prototype and we anonymize them upon request.
- The data stays in the EU and we delete it after the test.
- Corporate accounts with an administrator on your side.
When data cannot leave the company
The open model runs on your server or in your cloud in the EU. To choose from:
- White-tailed eagle
- Polish model, SpeakLeash and ACK Cyfronet AGH
- PLLUM
- Polish model, HIVE consortium commissioned by the Ministry of Digitization
- Qwen
- Alibaba
- GLM
- Zhipu AI
- DeepSeek
- DeepSeek
We select a model for the task and check it on your cases. We run it, tune it and maintain it.
Source: Bielik and PLLuM, ACK Cyfronet AGH ↗ · PLLuM, Ministry of Digitization ↗
Do you have a lot of your own data? We will fine-tune the small model.
- 1 Your examples
- 2 tuning
- 3 test on new cases
- 4 model at your place
When you have thousands of similar emails, documents or reports, we teach a small open model based on your examples. It works for you, responds quickly and costs less to maintain than a large model. We measure quality on issues that the model has not seen while learning.
Przemysław and Piotr did it in large programming and consulting companies. Artem did it before and does it today at Syntalith. It's about models on your own infrastructure, tuning and quality measurement.
You choose the protection level during the interview. We sign the data processing agreement before we see any data.Schedule a free analysis →
Price
Three packages. You know the price right away.
Team of up to 10 people
from €1,150
net, paid once · 2 weeks
Company, from 11 to 40 people
from €2,300
net, paid once · 3–4 weeks
Organization, over 40 people
from €4,600
net, paid once · 6–8 weeks
- The same packages once you have licenses. To start, an inspection of the space in 1 day is enough: €450 net.
From conversation to measurement
01 · 30 min
Free process review€0
An engineer picks 3 to 5 of your team's tasks with you. Written plan within 2 working days.
02 · 2 weeks
Rolloutfrom €1,150
Accounts and roles, rules for working with data, training on your team's tasks. You buy licences directly from the provider.
03 · after 30 days
Measurementincluded
We compare task times before and after. We name in writing the tasks worth handing to a system.
or call+48 888 78 48 78WhatsAppMon–Fri 9:00–17:00 CET
Questions
How do you choose an agency to implement AI?
Give every candidate the same brief and one real case. Assess the questions they ask, the demo of a hard case, the description of acceptance and the maintenance terms. Compare the price of the same result.
How much does an AI rollout by an agency cost?
At Syntalith an audit costs from €600 net, an automation from €3,000, an app €3,000 to €7,000, an agent €3,000 to €17,500 and training from €1,150. You count licences, the model, hosting and maintenance separately.
What should you watch for in an AI agency's portfolio?
Tell work for a client apart from a test on synthetic data. Both show skill, and only the first says something about a result in a company. The result of a mechanism test does not establish savings in your company.
Is an n8n agency enough?
Yes, when the process fits ready-made connections and the team can handle exceptions in its current tools. A review panel, roles and custom logic are separate items to price.
Does Syntalith hold partner status?
Yes. Syntalith holds OpenAI Select Partner status in the OpenAI Partner Network and is a member of Anthropic's Claude Partner Network. That tells you about the ecosystem, and a test on your process decides the choice.
