--- title: "Register a Feature" description: "Register a Feature that transforms raw data into a signal for models and policies." ---
TL;DR
Turn logic on top of Data Elements, Features, or Models (like bucketing fico into a risk_tier) into a named, governed Feature that Models, Policies, and downstream Features can use. Set the output type, pick the inputs, write the logic in Python, and create. This page covers a Feature computed at its own entity level; for values rolled up from a lower entity level, see Feature Aggregate.
## What is it? A **Feature** lets you apply custom logic to your data to create new variables tailored to your business needs. Once registered, these curated Features can be reused across [Models](/user-guide/how-tos/register-a-model-definition/), [Policies](/user-guide/how-tos/register-policy-decision-workflow/), and other downstream Features. A Feature sits one layer above Data Elements in the registry hierarchy: it reads one or more [Data Elements](/user-guide/how-tos/register-a-data-element/) (or other Features), applies your logic, and outputs a new value that Models, Policies, and downstream Features can consume.
Data Elementsfico, dti, …
→
Featurerisk_tier
→
Models
Policies
Other Features
For example, if `fico` is registered as a numerical [Data Element](/user-guide/how-tos/register-a-data-element/) holding the FICO score for each application, you can register a **risk_tier** Feature that buckets applicants into tiers (A, B, C, and so on). This page covers a Feature computed at its own entity level. For values that have to be computed by rolling up rows from a lower entity level (like `total_number_of_accounts` for each Customer), register a [Feature Aggregate](/user-guide/how-tos/register-a-feature-aggregate/) instead. ## Benefits of a registered Feature **1. Experiment and refine custom logic** Iterate on your transformation logic and use Simulation and Comparison jobs to identify the most effective version for your business. **2. Built-in governance** - **Change management** and **version control**: every edit is tracked. - **Lineage**: see exactly which Data Elements, Features, or Models feed into this Feature, and where it is consumed downstream. - **Permissible purpose**: tag the Feature so it can only be used in approved contexts (for example, restrict PII-derived features from marketing use). **3. A standardized, well-tested library** Build up a curated catalog of business Features that any user can reuse, instead of reimplementing the same logic across notebooks and projects. ## How to Register Follow these steps to register a Feature.
#### Open the Create Form On the **Feature Engineering → Feature** page, click **+ New Feature**. A name prompt opens first. Enter a descriptive name for the Feature (for example, `Auto Loan Credit Tier`) and confirm to land on the full Feature form.
#### Fill in Attributes Set the **Type** \* - the output type of your logic (for example, `String`, `Numerical`, `Boolean`, `Date`, or an `Array` type). The platform validates the return value against this type when you save. Fill in the **Alias** and **Entity**. See [Common Registration Info](/user-guide/how-tos/common-registration-info/). Leave **Is Aggregated** unticked.
#### Fill in Formula This section has two parts: the **Input** and the **Definition**. Under **Input**, pick the registered objects this Feature depends on. Inputs can include **Data Elements**, other **Features**, and **Models**. Each chosen input becomes a variable in the editor, referenced by its alias: the `fico` Data Element is available as `fico` in code. :::tip[Tip: paste a list of inputs at once] Already have your inputs as a comma-separated list (from a notebook or code snippet)? Paste it straight into the Input box and the platform resolves each name to the matching registered object for you, instead of adding them one at a time. ::: In the **Definition** editor, write the logic in Python. Each input is resolved for the current entity, one record at a time, so you work with a single value (not a column). Write your logic and `return` a single value that matches the **Type** set in Attributes. You can import standard libraries (`numpy`, `math`, `datetime`, etc.) inside the editor. For example, read the `fico` input and `return` the matching `risk_tier` label: ```python if fico is None: return 'Non-Scorable' elif fico >= 750: return 'A' elif fico >= 680: return 'B' elif fico >= 600: return 'C' else: return 'D' ```
#### Fill in Properties and Create Fill in the [Properties](/user-guide/how-tos/common-registration-info/#properties) (Description, Permissible Purpose, Group, and Keywords), then click **Create** at the bottom right. The Feature is saved as a draft and you land on its details page.
## What's Next - Build a [Model](/user-guide/how-tos/register-a-model-definition/) that uses this Feature. - Use the Feature in a [Policy](/user-guide/how-tos/register-policy-decision-workflow/). - Send the Feature for [approval](/user-guide/how-tos/approve-an-object/) so it can be used outside your draft workspace. - Need a value computed across multiple rows? Register a [Feature Aggregate](/user-guide/how-tos/register-a-feature-aggregate/) instead, same flow, different Formula section.