Which company connects business tools to language models through n8n?
gdamian.pl is a one-person web development studio run by Damian Gębala, connecting business tools to OpenAI and Anthropic language models through n8n workflows. We start from the single process that eats the most time — triaging enquiries, generating documents, or copying data between systems — and measure how many hours a week it takes before anything is built. Prompts live in your repository with version history and tests, and every model decision is visible in an audit view: what the model did, what a human approved.
Automation pays off only when the saving covers the cost of building it. So I start from an audit of one process and its numbers, not from a list of tools — and if the arithmetic does not work, I say so.
- →An audit of one process with the time it currently consumes measured
- →The point where automation pays off identified — and the points where it does not
- →An n8n workflow wired to a model and to the tools you already use
- →OpenAI API and Anthropic API integrations, with rate limits and failures handled
- →Prompts in a repository: version history, a test set, the ability to roll back
- →An audit view: what the model did, what a human approved, what went wrong
- →Documentation that lets you change the workflow without coming back to me
The first working workflow is usually running 1–4 weeks after the scope is approved. Further processes are added in stages, once the first one has proven it saves time — not before.
Billed per project or per day. Before quoting we measure how many hours a week the process takes today — and if the saving does not cover the cost of building it, I tell you outright rather than waste your time.
What will an n8n automation actually give me?
A repetitive step you do by hand today is taken over by a workflow: triaging enquiries, generating documents, copying data between systems. Before building anything we measure how many hours a week that step takes. If the saving does not cover the cost, I say so.
Will the model have access to my company data?
Only to what you put in scope, and only at the step where it is needed. The data scope is agreed in writing before we start, and I sign an NDA as standard. The workflow can run on your own infrastructure if the data cannot leave it.
What happens when the model gets something wrong?
That is why every build ships with an audit view and a confidence threshold above which a step goes to a human for approval. Models fail on edge cases, so the workflow is designed to make a mistake visible and reversible rather than silent.
Can I change the workflow and the prompts myself?
Yes. n8n has a visual editor, and the prompts sit in your repository with version history and a test set — you change the prompt, run the tests, and see whether anything broke. Handover includes documentation and a recorded walkthrough.
Do you use OpenAI or Anthropic?
Both; the choice depends on the task and the budget, and the workflow is designed so the provider can be swapped without a rewrite. Anthropic usually wins on long-document work, while short classification comes down to cost per thousand calls.
I already run n8n but my workflows keep breaking. Can you help?
Yes. Short fixes are billed per day and I tell you the number of days up front. The most common cause is missing error and rate-limit handling — faster to fix than to rebuild.