AI Training
AI Training for Companies
Syntalith provides closed AI training for companies and teams: one or two days, online or at the client’s office, from €1 200 net per day. Engineers work on participants’ tasks, data rules, output verification, and the competence measures required by Article 4 of the AI Act.
An engineer adapts the workshop to one process from your team. Participants practise setting context, protecting data, checking outputs, and deciding when a person must approve the result.
- A 30-minute call with the engineer who would lead the work.
- We review the process, its exceptions, and its current cost.
- Within 2 business days, we send a written summary.
In short
- Price
- from €1 200 / day nettypically €1 200–1 500 per day; quoted per scope after one call
- Format
- 1 or 2 days, online or on-sitea workshop group of up to ~15 people, split into several groups in a larger company
- Who teaches
- Engineers with systems in productionthe program comes from production deployments and work on your process
- For whom
- Boards and teams that use AI or are about to startfor teams that need a shared AI working standard on team tasks
Two formats
A team or one person?
Choose by who will work and what they bring to the session. The €1 200 daily floor belongs to the training line; the technical AI-Native course is quoted separately after a codebase call.
Closed team training
For a company that needs shared rules and exercises on the team's own tasks.
from €1 200/day net
Stay on the company training page →One-to-one AI training
For one person bringing documents, research, analysis or an operating process.
from €1 200/day net
See one-to-one training →Who leads the AI training, and what do we work on?
The workshop is led by an engineer who builds and maintains AI systems. Exercises use agreed team tasks and include data rules, verification, and an approval record.
| Criterion | A typical AI course | Syntalith training |
|---|---|---|
| Who teaches | A training facilitator | An engineer who ships AI systems in production |
| On what | Generic examples | Your defined process and data |
| What it teaches | Prompt tricks | Context, boundaries, escalation, trace |
| Compliance | Skipped | Operator duties under GDPR and the AI Act |
| After the session | Slides | Materials and a way of working that stay with you |
1:1 course participants
People we've trained one-on-one.
The reviews below come from five one-to-one sessions. The team format is an offer built on the same working standard.
Kuba Koziej CEO & co-founder of MoreGrowth, board member at Natu.Care
I went from barely shipping a frontend to building full applications with a backend and a database, and deploying them safely instead of hoping they would hold. We worked through the parts I kept tripping on: Docker, the coding agents and tools, and when it is worth reaching for skills, MCPs and connectors. It stayed practical the whole way, on what I actually wanted to build.
Hlib Utkin
Public administration
A large stream of documents passes my desk every week. I now review, draft and prepare them with Claude in a fraction of the time, and research that used to eat an afternoon takes minutes. It is wired into the services I already work in, so the documents reach me instead of me hunting for them.
Ivan Chepurin Senior Software Engineer, Immutable
I came in using coding agents ad hoc and left orchestrating them: loops, an agent kanban, a full development cycle that ships. What stayed with me is running context and cost on purpose, and building my own harnesses instead of waiting for a tool to ship one. It changed how I work day to day.
Oleksandr Usyk Co-founder & Art director, jakotako
I run a design studio, not an engineering team, and I still left with something working. We built an agent loop that pulls in leads and keeps an eye on the competition, and I put my own site together with Claude. It was hands-on, on my real work, not a talk about AI.
What will the team learn during AI training?
The team practises providing the right context, protecting data, checking results, and recording important decisions. Each module applies those habits to an agreed task.
Theory · the foundation
How the models work and where they fail, on neutral examples before we work on your tasks.
01Foundations · ModelAI foundations: how models work and where they fail
How language models produce answers, why plausible errors occur, and which outputs require verification before company data is introduced.
- 02Task context
Context, task, and shared work in the selected tool
How to give the model context and a task, how to run multi-hour work on one document, and when to start the conversation again.
Practice · the drills
Concrete drills on your process: tool, boundaries, escalation, trace.
03Documents · AnalysisAI tools in the team’s daily work
Documents, files, offers, analyses, correspondence: your team's everyday tasks moved onto the model, with verification of the result.
- 04Tools · Boundaries
Agentic workflow: prompt, automation, or agent
When a conversation with the model is enough, when a repeatable workflow is better, and when an agent should use tools, and where a human must approve the result.
- 05GDPR · Escalation
Data safety and oversight
Work aligned with company policy and GDPR: what may be entered, what must stay out, which data settings to enable, where the tool must stop, and when a person takes over.
- 06Trace · AI Act
Trace, quality, and AI Act duties
How to document important decisions, measure the quality of AI-assisted work, and support the AI-literacy measures required by Article 4.
Who is company AI training for?
The program starts with the participants' roles and one process they handle. Executives, operations teams, and developers receive different exercises.
Boards and decision-makers
You decide where AI enters the company and at what cost, and every vendor's offer sounds the same.
- Where AI and agents are an advantage, and where they are a risk
- How to read an AI vendor's offer and what to demand in the contract
- What the AI Act means for a company deploying AI
- A short session focused on operational decisions
Operations teams
Your people already reach for AI on the job, each in their own way and with no rules.
- Hands-on work in company-approved tools using your team’s tasks
- Context, boundaries, and catching the plausible-but-wrong answer
- What can go into the model and what can't, data-handling rules
- Materials and a working standard that stays with the team
Team leads
You have to roll AI out in the team so it sticks and doesn't fracture into 20 styles.
- How to spread AI work across the whole team
- Shared rules and a working standard across the whole team
- What stays in the team after the training, and how to keep it
- Keeping quality and a trace in daily work
Formats and prices
from €1 200
per day · net
How much does company AI training cost, and which format should we choose?
Training starts from €1 200 net per day. Choose one or two days, online or at the client's office; the final quote depends on scope, participant count and the number of groups.
| Format | For whom | Scope | Price |
|---|---|---|---|
| 1 day · introduction | Boards, decision-makers, teams starting with AI | One day on your own examples: what AI and agents do well, where the boundaries are, and which data-handling rules to adopt from tomorrow. | from €1 200 per day · net |
| 2 days · workshop on your process | Teams that will use AI daily | Day one: hands-on work with the team’s tasks. Day two: boundaries, exception handling, decision records, and a working standard the team keeps. | from €1 200 per day · net |
The day rate typically falls between €1 200 and €1 500 net; final quote per scope after one call. The program, attendance list and materials stay with you.
Closed team training or an open course: which should we choose?
We described above which group you fit; this is about readiness. If the second column describes your case, another provider usually fits better, and we'll say so before the quote.
Ready if
- You have a defined process on the table, beyond curiosity
- Your team is about to reach for AI on the job and needs boundaries first
- You care about a decision trace and the Art. 4 AI Act literacy duty
- The session should use your own data instead of generic examples
- Working with AI should become a habit for the whole team, beyond a couple of enthusiasts
Probably not if
- You want a trend lecture with no work on your own tasks
- You only need prompt tricks and a certificate for the wall
- You want cheap off-the-shelf training with no fit to your process
- You expect a guaranteed result or an official endorsement from the model vendor
- You are a technical founder who wants to build software with agents: see the AI-Native course
AI Act · Art. 4
Can AI training support the Article 4 AI Act duty?
Article 4 of the AI Act requires providers and deployers to take proportionate measures toward a sufficient level of AI literacy among people using AI systems on their behalf. It does not prescribe one course. A program, materials and attendance record can support the company's documentation, but they do not establish compliance automatically. This is not legal advice.
EU AI Act 2024/1689, Art. 4 (EUR-Lex) →How do you train employees so AI changes the way they work?
Start with one process, shared rules and exercises on the team's tasks. Training begins with a call about the group and its goal, then ends with materials and a recorded working standard.
- 01A call about the team, process and goal: we set the group, format and scope.
- 02You get a quote and a program tailored to your tasks and data.
- 03We run the training online or on site, on prepared examples.
- 04You keep the materials, attendance list, named certificates of completion and a working standard that helps document AI literacy under Art. 4.
What being in the Claude Partner Network means.
We say plainly what the status means, and what it doesn't.
What it isn't
- Partner status does not mean Anthropic endorses our services.
- Partner status is not a training-quality certificate or a guarantee of results.
- Syntalith prepares and delivers this training; it is not an official Anthropic course.
What it means
- We are in Anthropic's partner program for Claude.
- We use Claude in production, in the automations and agents we build for Polish companies.
- We teach from practice: how Claude behaves in deployed systems.
Training at the client's office
Eight cities, one closed format
The trainer travels to the client's office or runs the workshop online. Each city page states the delivery logistics and the limits of any local claim.
Training questions
How soon can you run it?
Usually within a few weeks of the first call: that's the time to set the group and scope and tailor the program to your tasks and data. We'll try to fit an urgent date if the calendar allows.How many people can take part?
There is no upper cap: as many people from the company take part as needed. One workshop group runs up to ~15 people, because above that a workshop stops being hands-on work and becomes a lecture. Larger teams are split into groups and priced in one offer.Do you run AI training at the client's office?
Yes. An operating workshop can run online or at the client's office. For executives we usually recommend online delivery, while the location for a team workshop depends on the tasks and rules for handling the material. Logistics and travel are included in the quote.Do we need Claude or ChatGPT accounts and licenses?
We settle participant access to Claude or ChatGPT during scoping: if the company doesn't have licenses yet, we'll advise which plan makes sense for the team. We don't sell licenses and don't require a purchase before the training.Who leads the AI training?
Because the program comes from deployments. The same people build and run AI systems working in Polish companies: in the training they show the working standard they use in production themselves.How does the team keep using the method after the workshop?
Day two turns the exercises into shared rules and a working standard based on the team's own tasks. Participants keep the materials and a list of agreed decisions.Do you train teams to use Claude, ChatGPT, Copilot and Gemini?
We work in the tools the team uses, most often Claude or ChatGPT. Copilot, Gemini or another environment can be included when it fits the tasks, data rules and company licenses. The method is not tied to one provider.Do you provide AI training for executives, managers, sales, HR, accounting and legal teams?
Yes, as a closed offer for one company. We fit the program to the roles and one defined process: executive decisions, sales work, HR documents, accounting analysis or legal-team tasks. We do not present these variants as previously delivered department-wide training.Do you offer one-to-one AI training?
Yes. We have a separate one-to-one format for one person with a specific task. See one-to-one AI training.
Next routes
Program, pricing and decision material
Tell us who we're training and we'll send you the program and the quote.
- A 30-minute call with the engineer who would lead the work.
- A review of the processes that cost you the most time and money.
- A written summary of what to automate first and the likely cost range.
Times are shown in your own time zone. We work with clients across time zones.
The scan is free and creates no obligation. If automation is unlikely to pay off, the written recommendation will say so.
from €1 200/day net · quoted per scope · taught by engineers with systems in production