Data
The same record in several places
A lead, invoice or ticket lands in a form, the CRM, an inbox and a spreadsheet. Every retype is a chance for error.
Hours a week spent copying.
Automation studio · n8n · human in the loop
A studio for process automation and AI
We automate the busywork so people can get back to work that matters. We design n8n workflows, AI agents and system integrations — always with a human in the loop and with care for the team’s well-being.
- Human in the loop in every implementation
- Less busywork, less team overload
- Official n8n ambassador in Poland
- Web formNew inquiry
- AI agentScores and enriches
- Sales rep approvesOne-click approval Human in the loop
- CRM + follow-upRecord, task, reminder
The machine handles the routine; a person makes the call.
- n8n
- OpenAI
- Claude
- Gemini
- ElevenLabs
- HubSpot
- Pipedrive
- Salesforce
- Google Workspace
- Microsoft 365
- Slack
- KSeF
- PostgreSQL
- REST API
Sound familiar?
Manual work that quietly costs you the most
Automation makes sense where a process is repetitive and delays hurt sales, customer service, finance or reporting.
Ownership
The process works as long as someone remembers
Follow-ups, payment reminders and customer replies depend on one person. They go on vacation — and the process stops.
Lost leads and late payments.
Reports
Reports are built by hand — and too late
Data is scattered, so the report has to be stitched together from many sources — usually when it’s already too late to act.
Decisions based on stale data.
Support
Experts answer simple questions
Customers ask about status, deadlines and documents. Specialists check several systems instead of handling the hard cases.
Slower response times.
Services
Five services. One studio, from audit to monitoring.
A simple offer: we start with an audit, implement automation and AI, and train your team. For larger changes, we add the human component.
01Start here
Process audit
An automation map: what to automate first, what to leave alone, and why.
Details 02AutomationProcess automation and n8n
n8n workflows and CRM, ERP and API integrations — no more retyping data by hand.
Details 03AIAI agents, chatbots and voicebots
Assistants that answer, analyze and prepare decisions — with a human in the loop.
Details 04TrainingAI literacy training
AI skills for teams and leadership — in response to Article 4 of the EU AI Act.
Details 05Human componentTeam change
Employee well-being, change communication and support for managers during AI adoption.
DetailsIllustrative example
The same lead-handling process — before and after automation
A typical change at a B2B services company. The process stays the same; the manual retyping and deadline-chasing go away.
Before automation
- An inquiry lands in a shared inbox and waits for someone to notice it.
- A sales rep creates the contact in the CRM by hand, copying details from the email.
- Priority is a gut call, and the follow-up goes into a calendar — or gets forgotten.
- The manager doesn’t know how many leads are waiting until they ask.
After automation
- The lead goes straight into the CRM, with company details filled in.
- AI scores fit and urgency; rules assign a sales rep.
- The task and follow-up are created automatically; no response triggers a reminder.
- The team and the manager see every lead’s status in real time.
How we work
First we clean up the process. Then we automate it.
We don’t start with a list of tools. We look for the slice of work that will deliver a measurable result, pin down the rules and exceptions, and only then build.
-
01
Audit
We pick the process with the most potential
The business goal, process owners, data sources and the points where delays or errors happen.
-
02
Design
We define the MVP and the rules
Data, statuses, integrations and test scenarios — plus the human-in-the-loop points where a person makes the decision.
-
03
Build
We build and test on real data
n8n workflows, integrations, an AI agent or a chatbot — tested against typical and edge-case scenarios.
-
04
Grow
We monitor, train and improve
Alerts, documentation, team training and a plan for the next improvements — automation isn’t a one-off.
In every implementation
A working process, not just “connected apps”
- DocumentationHow the logic, data and exceptions work, plus instructions for making changes.
- Monitoring and alertsA workflow error never fails silently — someone gets notified.
- Access controlSeparate permissions, secure keys and event logs.
- Handover to your teamTraining on how to use and extend the automation.
Our principles
Human in the loop and well-being: automation should take work off people, not replace their judgment
Three principles we apply in every implementation — from a simple workflow to an AI agent. How we design human in the loop.
Human in the loop
AI prepares, people decide
Before implementation, we agree what the system does on its own and what a person approves. Decisions about money, customers and people are approved by a human — with the full context of the case, in one click.
Well-being
Less busywork, less overload
We take the draining tasks off the calendar: retyping data, chasing deadlines, answering the same questions. For larger changes, we also check the team’s morale and workload — before and after implementation.
Transparency
No black boxes
The team knows what the automation does, why, and who owns it. Every implementation comes with documentation, logs and a simple way to report an error.
Human component
Technology is 80% of an implementation. The other 20% is employee well-being — and whether the change sticks.
AI rollouts rarely fail on technology. More often they fail on resistance, fear of job loss and team overload. That’s why, for larger changes, a psychologist joins the project to look after people’s well-being through the change — see how it works.
Technology · 80%
Automation and AI
Audit, architecture, n8n, AI agents, integrations, testing, monitoring and training. Led by Michał Kozak.
- solution design and build
- documentation and monitoring
- AI literacy training
People · 20%
Team change
Team sentiment assessment, change communication and support for managers. Led by a psychologist who works with us.
- anonymous sentiment assessment
- change communication
- support for managers, including around role changes
Use cases
Start with the problem, not the technology
Pick the area where leads, replies, documents or data get lost today. Each page shows what can be fixed and what implementation looks like.
When customers are waiting for answers
Customer service automation
Chatbots, voicebots, ticket routing and escalation to a person.
Learn moreWhen leads go cold after first contact
Sales automation
Leads straight to the CRM, qualification, routing and follow-ups that never rely on memory.
Learn moreWhen data is scattered
Marketing automation
Campaign contacts in your CRM, segments, messaging, AI-assisted content and reports.
Learn moreWhen invoices bounce between people
Invoice automation
Capture, OCR/AI extraction, validation, approvals, KSeF and export to accounting.
Learn moreWhen reports only happen on request
Reporting automation
KPIs from multiple systems, scheduled reports, dashboards, alerts and AI summaries.
Learn moreSelf-check — 30 seconds
Is your process a good fit for automation?
Check the statements that describe the process you have in mind. If it meets at least three, it’s a good candidate for a first implementation.
Studio
A small studio. Hands-on in every implementation.
Michał Kozak leads the technology; a psychologist who works with the studio leads the human component. Meet the studio.
Technology
Michał Kozak
Founder · automation and AI architect
Automation and AI architect, AI trainer, official n8n ambassador in Poland and author of the book “Inteligentna firma” (“The Intelligent Company”). Designs automations and AI agents that run in production — with documentation, monitoring and clear ownership.
- process audits and AI architecture
- n8n, AI agents and integrations
- AI training for teams
Human component
Collaborating psychologist
Supporting the team through change
A psychologist experienced in working with adults, following the principles of professional ethics and working under supervision. Joins projects where the change deeply affects the team, and is responsible for how people get through it.
- team sentiment assessment
- change communication
- support for managers
EU AI Act · Article 4
The AI literacy obligation may apply to your company too
Since February 2, 2025, companies that use AI systems — including popular assistants and chatbots — must take measures to build their people’s AI literacy. We run hands-on training tailored to each role and help you document those measures.
Where we work
We work across Poland
We work remotely, so location doesn’t limit our collaboration. See how we approach automation in Poland’s largest business hubs.
Mazowieckie
Warszawa
Centrale firm, finanse, usługi profesjonalne i centra usług wspólnych.
ViewMałopolskie
Kraków
Centra usług biznesowych, IT, turystyka i hotelarstwo, uczelnie.
ViewDolnośląskie
Wrocław
Przemysł i motoryzacja, logistyka, IT, centra usług i e-commerce.
ViewWielkopolskie
Poznań
Handel, dystrybucja, e-commerce i rodzinne firmy produkcyjne.
ViewPomorskie · Trójmiasto
Gdańsk
Porty, spedycja i logistyka morska, IT, produkcja i turystyka.
ViewŁódzkie
Łódź
Logistyka i magazyny, moda i e-commerce, centra usług, produkcja AGD.
ViewLubelskie
Lublin
IT, uczelnie, przemysł rolno-spożywczy, handel i sektor medyczny.
ViewBlog
Latest articles from the studio
Guides on n8n, AI agents and process automation — drawn from real implementations.
Example: automating invoices and document workflows
Retyping invoices by hand costs hours of work every month and risks accounting errors. Using an illustrative example, we show how automation — from OCR capture and coding to electronic approval — shortens an invoice’s path to posting, while the decision about the cost stays with a person.
Read article
Example: automating customer service tickets
Requests waiting in a single queue mean frustrated customers and missed deadlines. Using an illustrative example, we show how automation — from categorization and team assignment to SLA monitoring — speeds up customer service, while the reply stays with the agent.
Read article
Example: automating lead handling in a B2B company
Every hour of delay in contacting a lead lowers the chance of a sale. Using an illustrative example, we show how automation — from collecting inquiries and lead scoring to a first response within minutes — brings order to lead handling, and where the decision stays with the sales rep.
Read article
FAQ
Common questions before you automate
Don’t see your question? Email or call us — you’ll get a straight answer, not a sales pitch.
What does an automation studio do?
We design and implement business process automation: n8n workflows, system integrations, AI agents, chatbots and voicebots. We work like a design studio — small, hands-on in every project, from the audit through to monitoring.
Which process should we automate first?
One that is repetitive, frequent and measurable: lead handling, invoices, support tickets, reports or data moving between systems. We pick one slice (an MVP), ship it, measure the impact and only then expand the scope.
How long does the first automation take?
It depends on the number of systems, data quality and access. A simple workflow connecting two or three tools usually takes days; a process with an AI agent, multiple integrations and approvals takes a few weeks. We give you a specific timeline after a short process review.
How do we start implementing AI in our company?
With a problem, not a tool. We pick one process where AI will genuinely take work off people, agree where a human makes the call, and build a first version on real data. In parallel, it’s worth training the team in AI literacy — AI adoption succeeds when people understand how to use it.
Will automation work with our current CRM, ERP or spreadsheets?
In the vast majority of cases, yes. We connect systems through APIs, webhooks, n8n and ready-made integrations with sales, accounting, marketing and customer service tools. We check technical limitations and security concerns before we quote.
What does human in the loop mean in automation?
Human in the loop is the principle that automation and AI prepare the data, the analysis or a proposal, while a person makes any decision that affects money, customers or people. Before implementation, we agree for every step what the system does on its own, what it only prepares for approval, and what always stays with a human.
Will automation take jobs away from our people?
Our goal is to take work off people, not to replace them. We automate retyping data, chasing deadlines and answering the same questions — the work that wears people down and leads to overload. The team spends the time it gets back on harder cases, customer contact and growth. If the change affects headcount, we help you handle it fairly and respectfully.
What about data security and GDPR?
We design automations on the principle of least privilege: separate accounts and permissions, encrypted connections, event logs and — when needed — self-hosted n8n on your own infrastructure. In AI processes, we spell out which data goes to the model and where a human approves the decision.
Is AI training mandatory?
Article 4 of the EU AI Act has applied since February 2, 2025, and covers companies that use AI systems — including when employees use publicly available tools. A 2026 amendment (the Digital Omnibus) softened the wording, but did not remove the obligation to take measures that build AI literacy.
Why is there a psychologist on the project team?
Because AI rollouts stall on people more often than on technology: on job worries, overload and a lack of trust in decisions. For larger changes, we add a human component led by a psychologist who works with us. It’s an add-on, not a required part of every project.
Do you work with companies across Poland?
Yes. We work remotely with companies all over Poland — including Warsaw, Kraków, Wrocław, Poznań, Gdańsk, Łódź and Lublin. Workshops and sign-offs happen online, and when it makes sense, we meet on site.
Contact
Let’s talk about the process you want to lighten
Briefly describe what takes up most of your team’s time today. We’ll come back with a suggested next step — no strings attached.
Every inquiry is treated confidentially — including those about changes in your team.