First post in my blog, yeah!

Hi, it’s Nikolay Dyulgerov, or Nick for short.

I’ve been a software developer for the last almost 8 years, during which I’ve focused mostly on backend development, although I was a full-stack developer for the first half of my career so far.

I certainly have so much to learn and feel more curious than ever, therefore most of my posts here would probably be related to me trying to get a better grasp of some library, experimenting with some tool, or tinkering with some experimental software of my own.

I’d also like to refer to the history of this blog from time to time for my own calibration.

Posts about Elixir can be expected, as well as Elixir libraries, but I’m dabbling with Python, Go, Rust, even C a bit, so I might include some posts about my learning process, too. I might also write about my experience with LLMs and AI agents.

And who knows, I might find my love for Ruby again. Well, it’s not like I stopped enjoying it, but Elixir took the mantle of preferred language for me a few years ago.

In order to not let this post be just an introduction, let me share a few resources that made me even more excited about Elixir recently.

Resources

• Keynoyte: Type System and Elixir Updates + Extended Q&A - José Valim | ElixirConf EU 2025 - this is probably the biggest change coming to Elixir, and it’s been coming gradually over the last few years. However, it seems to be knocking on the door now. That’s the Elixir type system.

• Keynote: Rethinking Serverless with FLAME - Chris McCord | ElixirConf EU 2024 - slightly older talk, but there are newer talks for FLAME, too, and I expect there to be more in the near future. FLAME is kind of like Serverless on Elixir. It deploys containers of your app for more resource-intensive one-off jobs.

• The Next Evolution in ML Infrastructure — The NX Project - Elixir’s main contribution to the AI/ML space. Nx is something I haven’t used much yet, but it’s definitely on my radar, especially in the current climate.

• Introducing Tidewave Runtime Intelligence - Tidewave basically allows developers to their Phoenix applications to their AI agents with an MCP server that can be integrated in Phoenix