You can embed the function declaration followed by the function in the alias itself, like so:
alias my_alias='f() { do_stuff_with "$@" (arguments)" ...; }; f'
The benefit of this approach over just declaring the function by itself is that you can have a peace of mind that your function is not going to be overriden by some other script you're sourcing (or using .), which might use its own helper under the same name.
E.g., Suppose you have a script init-my-workspace.sh that you're calling like . init-my-workspace.sh or source init-my-workspace.sh whose purpose is to set or export a bunch of environment variables (e.g., JAVA_HOME, PYTHON_PATH etc.). If you happen to have a function my_alias inside there, as well, then you're out of luck as the latest function declaration withing the same shell instance wins.
Conversely, aliases have separate namespace and even in case of name clash, they are looked up first. Therefore, for customization relevant to interactive usage, you should only ever use aliases.
Finally, note that the practice of putting all the aliases in the same place (e.g., ~/.bash_aliases) enables you to easily spot any name clashes.
