Transformation Step

The Transformation step in RiskOS™ lets you manipulate data within a workflow using a set of predefined, stateless functions. Transformations do not maintain state; they operate only on the data provided as inputs at the time of execution.

For example, you can use the CONCAT function to combine firstName and surName into a single fullName variable. That new variable becomes immediately available to downstream steps, simplifying logic and reducing duplication.

How transformations work

Each transformation is a predefined, stateless operation: given the same inputs, it always produces the same output, and it keeps no memory between runs. A Transformation step has three parts:

Inputs — You supply each input as either a Value (a static string or number) or a Variable (a payload field or the output of an earlier step). The step passes inputs to the function in the order listed for that function, so order matters.
Output — The function returns a single typed value: a string, number, Boolean, array, or — for pass-through functions such as COPY — any type. RiskOS™ stores that value in the output field you name, which downstream steps reference as $outputFieldName.
Type handling — In general, if an input is missing or is not the type a function expects, the function returns a safe default rather than failing the workflow: string functions return an empty string, numeric functions return 0, and Boolean functions return false.

Placement in a workflow

A Transformation step can be placed anywhere in a workflow, depending on your logic.
Common patterns include:
- Right after the Input step to normalize incoming data.
- Before a Condition step to prepare data for evaluation.
- After an Enrichment step to parse out specific data.

Configure a Transformation step

Add the Transformation step to your workflow, then click the step to open the side panel.
Select a Function from the dropdown (see the Available functions section).

Provide input parameters for the function:
- Value: For static values (e.g., strings, numbers).
- Variable: For dynamic inputs, such as payload fields or outputs from prior steps.
Specify the output field name.
- The function’s result is stored in this field.
- Use it later with $[output_field_name].
- Example: If you name the output field fullName, reference it downstream as $fullName.

📘
Note:
You can customize the Transformation step’s label by editing the text field at the top of the side panel.

Available functions

📘
Note:
This catalog lists supported stateless functions. Check release notes regularly to confirm the latest set of functions available in your environment.

Each row lists a function's inputs, output type, how it works, and a worked example (shown as inputs → output).

String functions

Transformation	Input	Output	How it works	Example
`STARTSWITH`	String + substring	Boolean	Returns `true` if the string begins with the substring. Leading whitespace in the string is ignored.	`"hello world"`, `"hello"` → `true`
`ENDSWITH`	String + substring	Boolean	Returns `true` if the string ends with the substring. Trailing whitespace is ignored.	`"hello world"`, `"world"` → `true`
`LENGTH`	String	Numeric	Returns the number of characters in the string.	`"Socure"` → `6`
`CONTAINS`	String + substring	Boolean	Returns `true` if the substring appears anywhere in the string.	`"Socure"`, `"cur"` → `true`
`TO_LOWER_CASE`	String	String	Converts every character in the string to lowercase.	`"SOCURE"` → `"socure"`
`CONCAT`	Multiple strings	String	Joins the input strings, in order, into a single string.	`"123 "`, `"Main St"` → `"123 Main St"`
`SPLIT`	String + separator	Array	Splits the string into an array of substrings wherever the separator occurs.	`"a,b,c"`, `","` → `["a", "b", "c"]`
`FIRST_N`	String + numeric value	String	Returns the first n characters of the string.	`"Socure"`, `3` → `"Soc"`
`REVERSE`	String	String	Returns the string with its characters in reverse order.	`"abc"` → `"cba"`
`JARO_WINKLER_DISTANCE`	Two strings	Numeric	Returns a similarity score from `0` (no similarity) to `1` (identical) using the Jaro-Winkler algorithm. Does not normalize the inputs, so case and word order affect the score.	`"martha"`, `"marhta"` → score close to `1`
`SIAMESE_JARO_WINKLER_DISTANCE`	Two strings	Numeric	Same `0`–`1` Jaro-Winkler score, tuned for names. Normalizes the inputs, splits them into tokens, and sorts the tokens alphabetically before comparing, so word order does not matter.	`"Ryan Reisling"`, `"Reisling Ryan"` → score close to `1`
`SUBSTRING`	String + start + end positions	String	Returns the portion of the string between the start position (inclusive) and the end position (exclusive). Positions are zero-based.	`"Socure"`, `0`, `3` → `"Soc"`

Numerical functions

Transformation	Input	Output	How it works	Example
`DIVISION`	Two numbers	Numeric	Divides the first number by the second and returns the quotient. Returns `0` if the divisor is `0`.	`10`, `2` → `5`
`MULTIPLICATION`	Two numbers	Numeric	Returns the product of the two numbers.	`2`, `3` → `6`
`SUM`	One or more numbers	Numeric	Returns the total of the input numbers.	`2`, `3` → `5`
`MODULUS`	Two numbers	Numeric	Divides the first number by the second and returns the remainder.	`10`, `3` → `1`
`SUBTRACT`	Two numbers	Numeric	Subtracts the second number from the first and returns the difference.	`10`, `3` → `7`
`RANDOM_NUMBER`	Min + max values	Numeric	Returns a random number between the minimum and maximum values.	`1`, `10` → for example `7`

Array functions

Transformation	Input	Output	How it works	Example
`IS_IN_ARRAY_ANY`	String(s) + array	Boolean	Returns `true` if at least one of the input values is present in the array.	`"b"`, `["a", "b", "c"]` → `true`
`IS_IN_ARRAY_ALL`	String(s) + array	Boolean	Returns `true` only if every input value is present in the array.	`"a"`, `"d"`, `["a", "b", "c"]` → `false`
`JOIN`	Array	String	Joins the array elements into a single space-separated string.	`["John", "Smith"]` → `"John Smith"`
`ARRAY_LENGTH`	Array	Numeric	Returns the number of elements in the array.	`["a", "b", "c"]` → `3`
`COUNT_DISTINCT`	Array	Numeric	Returns the number of unique elements in the array.	`[1, 1, 2]` → `2`

Date functions

Transformation	Input	Output	How it works	Example
`DATE_DIFF_IN_DAYS`	Two dates (ISO) or date + “today”	Numeric	Returns the number of whole days between the two dates. Pass `"today"` as the second value to compare against the current date.	`"2026-01-01"`, `"2026-01-31"` → `30`
`CALCULATE_AGE`	Date of birth (ISO)	Numeric	Returns the age in whole years from the date of birth to today.	`"1990-06-15"` → age in years
`TIME_DIFF_IN_SECS`	Two timestamps (ISO)	Numeric	Returns the number of seconds between the two timestamps.	`"2026-01-01T00:00:00Z"`, `"2026-01-01T00:01:00Z"` → `60`

Date functions expect ISO 8601 dates (YYYY-MM-DD). A compact YYYYMMDD value such as "19900615" is not a valid ISO date, so CALCULATE_AGE cannot parse it and does not return a valid age. To convert a YYYYMMDD value inside the workflow, split it with SUBSTRING and rejoin it with CONCAT before calling CALCULATE_AGE: SUBSTRING($dob, 0, 4), SUBSTRING($dob, 4, 6), and SUBSTRING($dob, 6, 8) return the year, month, and day, then CONCAT($year, "-", $month, "-", $day) produces "1990-06-15".

Special functions

Transformation	Input	Output	How it works	Example
`GEO_DISTANCE_LAT_LONG`	Two lat/long coordinate pairs	Numeric	Returns the distance in km between two latitude/longitude coordinate pairs.	two coordinate pairs → distance in km
`EXTRACT_EMAIL_HANDLE`	Email address	String	Returns the domain portion of the address (everything after `@`).	`"[email protected]"` → `"example.com"`
`EXTRACT_EMAIL_USER`	Email address	String	Returns the local portion of the address (everything before `@`).	`"[email protected]"` → `"jane"`
`IS_NULL`	Any	Boolean	Returns `true` if the value is null or absent.	`null` → `true`
`COPY`	Any	Any	Copies the input value unchanged into the output field.	`"abc"` → `"abc"`
`DECLARE_CUSTOM_VARIABLE`	Any	Any	Stores a value in a named variable so later steps can reuse it (for example, as input to another step).	`"US"` → `"US"` (stored as a variable)
`INSPECT_VALUE`	Any	Any	Returns a variable's value together with its detected type. Use it to debug what a variable holds at a point in the workflow.	inspects a variable's value and type

Best practices

Place transformations where they add clarity or efficiency — e.g., normalize data early, simplify condition checks, or prep values before decisions.
Use clear names for output fields so they’re self-explanatory in downstream steps.
Validate transformation results with test payloads to avoid logic errors.