Run Elixir in the database using Dx - Arno Dirlam | Code BEAM Europe 2024

CodeSync · July 18, 2025, 1:00pm

This talk was recorded at Code BEAM Europe in October 2024. If you’re curious about our upcoming event, check https://codebebeameurope.com

Imagine you could write Elixir code as if all Ecto data is already loaded. With pattern matching to match associations; Enum functions to navigate associated data; accessing fields in associations without a second thought. Even using data from non-associated schemas without having to think about data loading at all, just using Enum functions to slice and dice the data as needed.

This is what the Dx library does. Just like Nx (numerical Elixir) translates Elixir code to run optimized for the GPU, Dx (data-driven Elixir) translates Elixir code to run optimized for the database. It translates your code either to Ecto queries or to a version with automatic data loading, optimized using batching and concurrency.

This makes for clear and easy-to-read Elixir code that’s structured in modules and functions as usual, without any need to implement data loading. The code is easier to maintain and refactor, compared to implementations with Ecto queries and/or preloading, yet efficient.

Comments welcome! View the code-sync tag for more Code Sync talks!

lud · July 18, 2025, 1:48pm

This is so cool!