I went to uni

No images? Click here

Goatmire Elixir 2026 website

28th of Sep - 2nd of Oct. This year. Varberg, Sweden.

The legends were true. A second coming of goat. Those who dare speak will find our Call for Talks. It is open! Yet Welcoming! Onward!

Heed the Call!

Seriously though. We want speakers of all experience levels from all parts of the world. Take a look. The previous conference was a blast.

 

Planting seeds of community

Thanks to an invitation by Johan Mattisson and an open-minded project group me and Alvise Susmel (aka Poeticoding) went to Växjö to speak and run a workshop at Linnéuniversitetet.

As part of it Johan pulled together the local Functional Programming Meetup so we presented there and then on the day after we were one of the workshops wrapping up a 3-year IoT and HPC project. They were lovely hosts. The attendees were curious and patient.

I have a number of learnings to file away for how I run this type of workshop. We were doing ex_nvr which gets you IoT, Elixir, Media processing and with some tweaks ML inference. And the stack of dependencies required for all of that is actually a bit much. I think people came away curious, thankfully. But we've already more or less settled on a VM-based setup if we do this again.

It was a a lot of fun. We also got to visit a local embedded + Elixir company. Not on Nerves, they are on Nvidia hardware, lots of context to be had there about the Jetsons.

Not only do I hope we made some people curious to try Elixir. I hope we made some professors curious to integrate it in their work and I'll try too keep those channels open.

If you university has a slice of budget left for a guest lecture or workshop and you'd like to see more Elixir around your seat of learning, do reach out. It is a nice way to build bridges and depending on the topics of the particular project or circumstance I know a lot of people I could rope into joining me. Watching Alvise talk to another ML-practicioner is like watching something snap to grid and lock in. From the challenges of sourcing labeled data for a new model, vector database choices and PyTorch CPU/GPU latency concerns. The reality is much more complex than the learning examples.

I had a great time. No recordings. A lot of what we worked on for this goes into my ElixirConf EU talk though so keep an eye out.

Now I get to say Arrivederci and make a rude gesture to my italian friends and go off towards the train station.

I'm here on email or on Bluesky and the Fediverse if you want to reach me :)

Thank you for reading. I appreciate you.

 
 

This is an email from Underjord, a swedish consultancy run by Lars Wikman.

Everything else is found at underjord.io

You signed up for this newsletter and confirmed the subscription. If you want to stop receiving it. Just use the link below.

Preferences  |  Unsubscribe