n8n: Automation Without Code. For Teams That Want to Deliver.
n8n is an open-source automation tool that connects systems, automates workflows and integrates AI models, without every connection having to become a development project.
Automation that stays in your hands.
n8n (pronounced ān-eight-nā) is an open-source workflow automation tool that connects systems through a visual interface. Instead of writing code, you build workflows by drag-and-drop: when event X happens, do Y with the data, then send it to Z.
The key difference from tools like Zapier or Make: n8n can be self-hosted. Data never leaves your server if you do not want it to. And those who need code can add it. n8n is not either-or, it is both.
With over 400 integrations, from HubSpot to Slack to PostgreSQL, and native AI connectivity to models like GPT or Claude, n8n is currently one of the most powerful tools for teams that want to automate processes without needing a development team.
Three things n8n is particularly strong at.
n8n is versatile. Here are the three areas where it makes the biggest difference.
Connect systems without API knowledge
n8n has ready-made nodes for most common tools. HubSpot, Slack, Google Sheets, Airtable, Notion, PostgreSQL, HTTP requests. Those who want to connect systems without a native integration can use the HTTP node to reach any REST API.
The result: Integrations that previously required developer time get built in hours.
Integrate AI models into workflows
n8n has native nodes for OpenAI, Anthropic Claude, Google Gemini and others. Summarise documents, classify emails, extract data from PDFs, generate responses: AI becomes a step in the workflow, not a separate project.
The result: AI automation without your own infrastructure and without prompt engineering at the developer level.
Self-host and keep your data
n8n runs on any server, any VM, any container. Docker image, own VPS, internal infrastructure: data never leaves your environment. For companies with data protection requirements or sensitive processes, this is often the deciding factor.
The result: Automation without dependency on a cloud provider and without data protection compromises.
On-premise or n8n Cloud: what costs what?
n8n Cloud (hosted by n8n): Starter from around ā¬20 per month, Pro from around ā¬50. Limited by workflow executions and active workflows. A good starting point for smaller teams without their own infrastructure.
Self-hosted (on-premise): The Community Edition is free. Infrastructure costs on top: a small VPS for simple workflows runs at ā¬5-20 per month. Those who need enterprise features like SSO, advanced user rights and support pay for the Enterprise licence, prices on request.
The key point: those who self-host pay no usage-based fees. A thousand workflow executions per day costs the same as ten. This makes self-hosting significantly cheaper at higher volumes.
Where and how does n8n run best?
n8n can run on almost any infrastructure. The most common options:
Own VPS (e.g. Hetzner, DigitalOcean): Simplest self-hosting option. Docker Compose, done. Sufficient for most small to medium workloads. Cost: ā¬5-20 per month.
Own VM in-house: For companies that want to keep their data completely internal. Full control, full responsibility for updates and reliability.
Container / Kubernetes: For production environments with higher requirements for scalability and reliability. More complex to set up, but more robust in operation.
What applies to all options: n8n needs a Postgres database for production use, and a proper backup strategy for workflows and credentials.
Why self-hosting is the only option for many companies.
As soon as workflows process personal data, customer data from the CRM, email content, invoice data, the question arises of where that data goes.
With n8n Cloud, data runs through n8n servers in the EU. That is sufficient for many, but not for everyone. Companies with strict internal policies, healthcare, finance or conservative data protection officers often require self-hosting.
The advantage: with self-hosting, the processed data never leaves your infrastructure. n8n itself does not see it. Your workflows, your credentials, your data.
What is often overlooked: AI models integrated into n8n workflows also process the data you send them. Those who call OpenAI or Claude via n8n are sending data to those providers. For sensitive data, this requires either locally hosted models or verified data processing agreements with the AI providers.
From manual export to automated pipeline.
A team was manually exporting data from their CRM every day, cleaning it in Excel and uploading it to the reporting tool. Two hours of work per day, error-prone, and when the person responsible was sick the reporting came to a standstill.
With n8n the entire process was automated: CRM data is fetched hourly, filtered, transformed and written directly to the reporting tool. The manual step is gone.
- Daily manual export replaced by automated workflow
- Error rate from manual processing reduced to zero
- Workflow runs hourly instead of daily, data is more current
- Two hours of working time freed up per day
- Self-hosted on own VPS, data never leaves the company
- Extended with an AI categorisation step in week two
INTRODUCE N8N WITHOUT IT BECOMING A PROJECT.
If you want to use n8n but are not sure where to start, which hosting option fits or how to integrate AI sensibly: talk to us.