How do I use Tines Formulas effectively for data transformation?

Tines Stories: Ask + Answer
1

I’ve been getting by with basic data referencing in Tines, but I know the formula system is much more powerful than I’m currently using it. I’d love to understand how to do real data transformation, things like reshaping objects, filtering arrays, working with dates, and handling missing values gracefully. Where do people start when levelling up their formula knowledge?

2

Sounds like you’re in a good spot to level up, and the good news is the formula system is pretty approachable once you know where to look.

One thing to know first

In the action builder, each field can be set to different types (formula, text, etc.) using the controls on the right.

data type switcher
data type switcher595×166 6.05 KB

Formula fields return typed data like objects and arrays. Text fields always produce strings. If you’re passing structured data to the next action, use a formula field.

The functions that’ll get you furthest

You don’t need to learn everything at once. A few functions cover most real-world transformation:

  • WHERE() filters an array of objects by a value, e.g. WHERE(alerts, "severity", "high") gives you only the high-severity ones.
  • MAP() extracts a single field from each object in an array, e.g. MAP(users, "email") gives you a list of emails.
  • DEFAULT() provides a fallback when something is null or empty, so you don’t hit errors downstream.
  • MERGE() and OBJECT() let you reshape and build objects from scratch.
  • DATE() and DATE_DIFF() handle formatting and comparing timestamps.

The formulas documentation has a full reference with examples for each of these.

Tip to avoid messy nesting

When you’re chaining a few transformations together, the pipeline operator (|>) keeps things readable. Instead of nesting functions inside each other, you read left to right:

WHERE(data, "active", TRUE) |> MAP(%, "name") |> SORT(%) |> JOIN(%, ", ")

You can also stack them vertically, which is easier to scan when things get longer:

WHERE(data, "active", TRUE)
|> MAP(%, "name")
|> SORT(%)
|> JOIN(%, ", ")

Where to go from here

Tines University has courses that walk through data transformation in context, so you’re working with realistic scenarios rather than isolated examples. The Story Library is also worth browsing since many published stories use formulas in interesting ways you can pull apart and learn from.

Honestly, the fastest way to get comfortable is to take a story where you’re passing raw API responses between actions and start reshaping at each step. It clicks quickly once you start experimenting.