We've all worked on projects that are big, tangled, and hard to change with confidence. These projects often have you source diving all throughout the codebase, always leading you deeper and deeper. Sometimes you never find your way out! These troubles can be caused by false layers, bad abstractions, and excessive nesting! Testing becomes harder once you introduce a side effect or access global state deep within the stack trace, or need access to an input provided at the boundary. We can avoid these problems and produce fast, testable, and modifiable software by embracing functional programming and removing waste. Together, let's explore how we can writing a non-trivial Phoenix application exercises these concepts! We'll see how to get a 100% `async: true` test suite, easily add new features, and access remote systems (DB, APIs, Queues). All without tearing our hair out.

Elixir is a language that has unique features that make it ideal for GenAI. But, the fact that it is still a relatively unused language means that there are challenges in using GenAI solutions for Elixir. I'll be discussing how Expert VibeCoding can shift the equation.

Tired of feeling boxed in by PaaS vendors? So were we. In this talk, we'll walk through how we deployed a production Elixir/Phoenix system to Google Cloud using Managed Instance Groups (MIGs). No Docker. No Kubernetes. We'll show how this setup gave us better performance, tighter control, and a smooth CI/CD pipeline powered by tools you probably already use today, like GitHub Actions. You'll learn how to scale your app with load balancing, autoscaling, secrets management (via Doppler or GCP Secret Manager), and VM-level access for fast debugging. If you've ever felt like deploying Elixir is hard, or that you must use a platform-as-a-service, this talk will challenge that assumption and give you a concrete alternative being used in production today.

Dive into the technical aspects of how Elixir's concurrency model helps with parsing and cross-referencing large genetic datasets against ClinVar

When you call `Enum.map(1..10, fn x -> x + 1 end)`, do you know what's happening behind the scenes? `Enumerable` is perhaps the hardest to grok of Elixir's built-in protocols, but it's a critical part of the language. The `Enum` module, the `Stream` module, `for` comprehensions, and the `in` operator are all powered by `Enumerable`. This talk covers the `Enumerable` protocol in depth, breaking it into more understandable chunks. We will look at implementations for well-known data structures as well as the functions that build on `Enumerable`, graphically map how they interact, discuss streams, and discuss the strengths and weaknesses of the protocol.

The Elixir ecosystem provides exceptional primitives for concurrent programming, yet there exists a gap between basic task execution and full-featured job processing systems. Imagine you could launch hundreds—or even thousands—of little helpers in your Elixir app, each working on their own piece of a big problem, and you could watch them in real time: who's running, who's finished, who's stuck, and what they've accomplished. In this talk, we'll peek under the hood of Elixir's BEAM virtual machine to see how Tasks and Streams really work—how they're created, how they communicate, and why they almost never crash your app. We'll see why sometimes you want more control and visibility than Elixir's built-in tools provide, but without the overhead of a full job system like Oban. You'll leave with practical ideas for making your async workflows safer, smarter, and easier to debug—without making your codebase a tangled mess. Let's make concurrency not just powerful, but transparent and fun!

Elixir is known by its community for having best-in-class tools, so what else is needed in order to grow adoption? Great frameworks, libraries, and tooling create a strong foundation for opportunity, and in this talk I will be presenting on how Elixir programmers can use that foundation to further the adoption of the language and even start new businesses for themselves.