n8n, Forgejo and Audiobookshelf: homelab workflows

Explore Kubernetes homelab workflows, including hourly AI reviews of update pull requests and podcast transcription with n8n and Audiobookshelf.

Player not loading? Watch on YouTube

The speaker presents five applications and workflows he uses on his Kubernetes homelab. This is a tour of his setup rather than a step-by-step installation guide. The AI examples focus on automation and listening to research.

Forgejo hosts his repositories and CI/CD workflows. He describes a blog release process that builds container images and stores them in his own registry. He chose Forgejo over GitLab, which he considers too resource intensive for his needs, and preferred its open source approach to Gitea.

In n8n, an hourly workflow checks application update pull requests. The speaker says AI assesses which updates can merge automatically and leaves a report for later review. His custom TTS application turns pasted essays into audio that he can access on his phone during walks. He does not name its speech engine.

For health tracking, he collects Whoop and Oura Ring data in Postgres databases, then uses Grafana to inspect sleep and HRV trends and configure alerts. He describes running databases on Kubernetes as advanced homelab work.

Audiobookshelf serves his audiobook collection and downloads podcast feeds to disk. He connects those files to n8n workflows for transcription, summaries and notifications about whether an episode might be worth listening to.