Kestra vs n8n: Docker setup and workflow comparison

The presenter builds the same AI news digest in Kestra and n8n, running both locally in Docker on the same machine. The workflow fetches Hacker News stories, filters for AI topics and produces a scheduled digest. Both implementations succeed; the supplied metadata reports a 4.54-second Kestra run.

Kestra stores the workflow in a YAML file, with Python filtering and an 8 a.m. schedule. The tutorial covers launching it with a Docker command, opening localhost, pasting the workflow and changing its keywords. A filter mistakenly admits a cocktail story because its text contains the letters AI. After fixing the filter, the presenter demonstrates saved revisions and a side-by-side diff.

The n8n rebuild uses six connected nodes and JavaScript code. The presenter praises its visual editor and confirms that the self-hosted community edition runs the digest for free. The execution allowance discussed earlier belongs to n8n's cloud Starter plan, rather than this local test.

Licensing drives much of the comparison. The presenter contrasts Kestra's Apache-2.0 license with n8n's internal-use restrictions and argues that selling hosted automation services needs closer license review. Kestra's trade-offs include a more technical interface, enterprise-only SSO and role controls, and a Java engine that the presenter says needs a couple of gigabytes of RAM.