[oskapt]

I run Huginn in a Kubernetes cluster with most of its heavy lifting done by HTTP calls to OpenFaaS running in the same cluster. This let me write complex logic in Python but have Huginn handle the execution, triggers, filtering, and delivery of the data. Among other things I use it to scrape Twitter for interesting Kubernetes news and follow the retweets back to people who link to their own websites so that our editorial department can vet them for invitation into our writing program. Any particular author/account only appears once. It generates around 40 new candidates each day. I also use it to scrape the Chilean public alert website (onemi.cl) for alerts in my region ( forest fire, major weather), run those through the translate API and then send the English translations to my and my wife’s phones via Pushover.

These are things that I want to control and customize, and I vastly prefer running my own systems over trusting any third party.