I've mentioned CareerCupid [0] a few times, a website where people can find out about other potential co-workers and companies that share their values and can be used a some form of screening system without forcing people to apply to a specific position.

To help overcome the chicken-and-egg issue, there is something a bit novel: the website integrates with ActivityPub, so you can follow @thecupid@cupid.careers (this is not a Mastodon bot, but a fully-fledged application built on my Django ActivityPub Toolkit [1]), I've made a quick demo [2] about on the last FediForum to show how we can build applications to leverage the existing social graph instead of trying to make the Fediverse grow by re-implementing federated versions of the traditional social networks.

When I first launched this, I made a post where I offered a discount for the first companies joining, but it didn't get a lot of visibility. Now that we got to the first of the month day and the "Who is hiring?" threads are upon us, I thought it would be a good time to say that the company pages and job listings are also integrated with ActivityPub and I'll be offering 3 months completely free for the first 100 people that signup (code HN2025)

If you want to try it out, you need to create an account and create a "company page" (https://cupid.careers/company/new), after creating the basic page, you will see the option to complete the subscription. Use the code and you'll be good to go.

----

It works in a similar way as OkCupid used to work: people answer questions and they indicate (a) how they would like other people to respond and (b) how important their answer is to you. By comparing your answers against everyone else's, we can create a "compatibility score" that can be a measure of how well you'd get along with a potential co-worker.

Another thing that I've been working on is what I am calling a "Skill Tree". My plan is to build make it easier for both hiring manager and employees to see what parts of the tech stack are really important vs what can be learned on the job or "adapted" by someone with experience in a related skill. When defining the skills, the same matching algorithm can be applied to score the technical fit.

For a (contrived) example: the expectations around "experienced with Python and PostgreSQL" can be quite different if we are talking about a data analyst, or a web backend developer, or a DBA that will need to write PL/Python routinely. So the idea here is to have this "Skill Tree" where there are many different root nodes (Software Development, Management, UI/UX, System Administrations) and let these skills branch out from these root nodes to make the skill requirements as broad (Software Development -> Javascript) or specific (Software Development -> Javascript -> React -> React Native, or Operations -> Continuous Delivery -> Jenkins) as needed. This

There are some other ideas that I'm considering, like creating a database of issues in open source projects that could be used as part of the skill assessment, a more reliable "vouching" systems where professionals can look at listings and make a referral even when they are not working at the company. But to do any of that, first I'd like to grow the user base and get more feedback.

[0]: https://cupid.careers

[1]: https://activitypub.mushroomlabs.com

[2]: https://spectra.video/w/nibHzQxR26zQX1eYkFgygL?start=2m44s