An additional option for some cases is Ecto.Query.select_merge which can populate virtual fields on your schema directly at read-time if the desired content can be calculated via SQL with fragment. This is something that you should measure the performance impact of but is generally within tolerances for queries that don’t produce an enormous number of rows.

This was very helpful to me @shanesveller, select_merge with a fragment is exactly what I wanted!

Took a little fumbling around to figure out exactly what to do, since none of the docs nor linked docs describe using fragments this way, so for reference:

defmodule Answer do
  use Ecto.Schema

  schema "answer" do
    field :word, :string
    field :length, :integer, virtual: true
  end
end

Repo.insert(%Answer{word: "SWANS"})

Repo.one(
  from answer in Answer,
  select_merge: %{length: fragment("length(word)")}
)
#=> %Answer{word: "SWANS", length: 5}

Tangentially, when posting code, please always use text instead of images. Text allows those who are helping to easily make suggested edits without retyping everything, and it’s also more accessible across a variety of screen sizes and resolutions.