Automation Made Easy - n8n
My current favorite tool for personal workflow automation
n8n is a powerful workflow automation tool that I've been using to connect various services and automate repetitive tasks. It is like a self-hosted alternative to Zapier or Make.
Why n8n?
- Self-hosted — Your data stays on your own infrastructure
- Open source — Free to use and customize
- Visual workflow builder — Easy to create complex automations
- Extensive integrations — Hundreds of built-in nodes for popular services
It's "low-code", meaning you can do an awful lot with just their workflow editor, but you can also include nodes that execute custom code in javascript or python. I've found that to be easier for some data manipulation tasks, while keeping the larger process as a visual flow.
Workflows are started by "triggers", like a schedule, manual execution, or incoming webhook. Then you can see how each node inside the workflow is executed, and the inputs and outputs at each step. It makes for a pretty powerful way to build and debug complex automations.
And of course, one of it's biggest draws is the "AI nodes" that let you integrate large language models directly into your workflows. You can provide them with prompts, input data, tools, and even structured output expectations.
My Setup
I run n8n on a small home server using Docker. This gives me full control over my automations without relying on cloud services. But it can be run on almost any machine with docker. You could run it directly on your personal machine or even a Raspberry Pi. And they offer professional cloud-hosted services if you need large-scale work.
I've got a docker-compose file which runs n8n and a couple of hand-made custom applications that I only use for it's interactions. I've mounted a couple of key folders from the host filesystem to allow it to interact with the files I need it to, but not let it run rampant across my whole server.
Later on I might write more about some of the specific workflows I've built, showing how I've used n8n to help me consistently practice some good habits, and reduce time wasted.
In the meantime, I encourage you to try it out youself! [Their documentation has a great docker-compose quickstart to get you going.]