n8n

Part of our topic guide on AI Skills for Business.

n8n (pronounced "n-eight-n", short for "nodemation") is a source-available workflow automation tool, published under a fair-code licence. It lets you connect different apps, databases and services together so that data moves and tasks run automatically, without someone doing it by hand each time.

You build a workflow visually, as a chain of connected "nodes" on a canvas — one node might watch for a new row in a spreadsheet, the next might clean the data, and the last might post it into Slack or a CRM. Because the source is available and you are free to self-host, you can run it yourself on your own servers or use the hosted cloud version, and you can extend it with custom code when a ready-made connector doesn't exist.

Why it matters

Most data and AI work isn't glamorous model-building — it's the unglamorous plumbing of getting data from one system to another reliably. A tool like n8n sits squarely in that space, and it's become popular precisely because so many teams are drowning in manual, repetitive handoffs between tools that were never built to talk to each other.

Our view is that the tool is never the hard part — knowing which process is worth automating, how to design it so it doesn't break, and how to check its output is trustworthy takes real capability, not just access to the software. Most AI and automation initiatives stall on exactly this: the demand side (people who can use the tool well), not the supply side (the tool itself). Learning to use something like n8n well is a good example of building that muscle.

How it works

• Trigger — a workflow starts on a schedule, a webhook, or an event in another app (a new email, a form submission, a database change).
• Nodes — each step in the chain is a node: fetch data, transform it, filter it, branch on a condition, or send it somewhere else.
• Connectors — pre-built nodes exist for hundreds of common tools (spreadsheets, databases, email, chat apps, CRMs); a code node covers anything bespoke.
• Execution — the workflow runs automatically each time its trigger fires, and you can inspect every run to see exactly what data passed through and where, which makes faults easy to trace.
• Deployment — self-host it for full control over your data, or use the hosted cloud version if you'd rather not manage the infrastructure.