A schema acts as a template for [creating and organizing records](/docs/assets/brickworks/quick-start/creating-a-record), ensuring consistency and structure across your data sets. In this article, you will walk through the process of creating a schema (**Products**) that includes the following fields: - the name of the item - the ID of the item - information about the item - recommendations which will return similar items to the defined item The details of the item will be provided while [creating a record](/docs/assets/brickworks/quick-start/creating-a-record). This example demonstrates how Synerise objects, specifically a catalog and an AI recommendation, can be utilized within a schema. While [creating a record](/docs/assets/brickworks/quick-start/creating-a-record) you will specify the item and will be able to preview recommendation results.
Before creating a schema for your business scenario, we recommend you become familiar [Limits and constraints](/docs/assets/brickworks/limits) as it may affect implementation of the scenario.
The example schema configuration
The example schema configuration
## Procedure --- 1. Go to Data Modeling Hub icon **Data Modeling Hub > Schemas > New schema**. 2. Select one of the following schema options: - **Managed schema** - **Simple schema** - **Singleton schema** See the ["Schema types" section](/docs/assets/brickworks/quick-start#schema-types) for an explanation for each type. 3. On the pop-up, provide the following data: - In **Display name**, enter `Products` In this field, you must provide the schema name; value in this field is displayed on the list of schemas only (Data Modeling Hub icon **Data Modeling Hub > Schemas**). - In **API name**, the value is pre-filled (`products`), if you want to change it, provide your own API name variant. The value in this field is the unique identifier used to reference this schema in API requests. - In **Description**, describe the schema’s purpose to help workspace users understand its function, for example, `Products with Synerise recommendations` This description will display on the list of schemas only. 4. Confirm by clicking **Apply**. ### Adding schema fields In this part of the process, you will add [fields](/docs/assets/brickworks/schema-field-types) to the schema. The schema will contain various field types: [strings](/docs/assets/brickworks/schema-field-types#string) and [Synerise objects](/docs/assets/brickworks/synerise-objects) (a [catalog](/docs/assets/catalogs) and [AI recommendation based on the similar item model](/docs/ai-hub/recommendations-v2/recommendation-types#similar-items)). This procedure focuses on the configuration of the fields. Later on, you will fill out the schema fields with the actual values, which is described in the [Creating a record](/docs/assets/brickworks/quick-start/creating-a-record) article.
A blank configuration form of a schema
A blank configuration form of a schema; that's how the starting point for creating a schema looks like
#### Field with the product name Add the [string field](/docs/assets/brickworks/schema-field-types#string) which will contain the product name. 1. Click **Add new field**. 2. From the dropdown list, select **String**. 3. Fill out the following fields: - **Display name**, enter the title of the field (`Product name`) which will display in the [record](/docs/assets/brickworks/core-concepts#record). - **Description**, optionally provide a purpose of the field to help other workspace coworkers understand the purpose of the field. 4. Select the **Required field** checkbox. 5. Confirm by clicking **Apply**. #### Field with the product ID Add the [string field](/docs/assets/brickworks/schema-field-types#string) which will contain the product identifier. 1. Click **Add new field**. 2. From the dropdown list, select **String**. 3. Fill out the following fields: - **Display name**, enter the title of the field (`Product ID`) which will display in the [record](/docs/assets/brickworks/core-concepts#record). - **Description**, optionally provide a purpose of the field to help other workspace coworkers understand the purpose of the field. 4. Select the **Required field** checkbox. 5. Select the **Unique value** checkbox. 5. Confirm by clicking **Apply**. #### Field with product information Add the [catalog field](/docs/assets/brickworks/synerise-objects#catalog) - it will let you select a catalog which will be the source of product information.
When creating a catalog, you select a primary key column whose values serve as unique identifiers for the catalog entries. In the [catalog field](/docs/assets/brickworks/synerise-objects#catalog) settings, you can specify how to provide the primary key value: either statically (by manually entering the value) or dynamically, by using [Brickwork Jinjava `{{ context }}` insert](/docs/assets/brickworks/brickworks-jinjava-inserts). In the scenario described in this article, we will use a static primary key value.
1. Click **Add new field**. 2. From the dropdown list, select **Catalog**. 3. Fill out the following fields: - **Display name**, enter the title of the field (`Product details`) which will display in the [record](/docs/assets/brickworks/core-concepts#record). - **Description**, optionally provide a purpose of the field to help other workspace coworkers understand the purpose of the field. 4. From the **Catalog** dropdown list, select the catalog in which you store the item feed. 5. Confirm by clicking **Apply**. #### Field with recommendations Add the [AI recommendation field](/docs/assets/brickworks/synerise-objects#ai-recommendation) and select the AI recommendation campaign whose results will be returned in the [record](/docs/assets/brickworks/core-concepts#record).
Some recommendation models based on which you can create [AI recommendation campaigns require item context](/docs/ai-hub/recommendations-v2/recommendation-types#recommendation-model-summary) based on which the results are generated. In the [AI recommendation field type](/docs/assets/brickworks/synerise-objects#ai-recommendation) settings, you can specify how to provide the item context (item identifier): either manually (by entering the item identifier) or dynamically by using [Brickwork Jinjava `{{ context }}` insert](/docs/assets/brickworks/brickworks-jinjava-inserts). In the scenario described in this article, we will use a static item context value.
1. Click **Add new field**. 2. From the dropdown list, select **Catalog**. 3. Fill out the following fields: - **Display name**, enter the title of the field (`Products`) which will display in the [record](/docs/assets/brickworks/core-concepts#record). - **Description**, optionally provide a purpose of the field to help other workspace coworkers understand the purpose of the field. 4. From the **Recommendation** dropdown list, select the recommendation whose results will be displayed in the [record](/docs/assets/brickworks/core-concepts#record) 5. Confirm by clicking **Apply**. ### Defining schema audience Defining [an audience for a schema](/docs/assets/brickworks/quick-start#schema-and-record-audience) ensures that the schema’s structure and data apply meaningfully and securely to the intended group of users.
This part of the process is optional. If skipped, the schema audience defaults to everyone.
1. Click the **Audience & Settings** tab. **Result**:
The Audience & Settings section
The Audience & Settings section
2. On the **Audience** section, click **Define**. 3. Choose the schema recipients: - **Everyone** - Everyone you have in [Profiles](/docs/crm). - **Segmentations** - Profiles in selected [segmentations](/docs/analytics/segmentations) will be schema recipients. 1. Click **Segmentations**. 2. On the pop-up, select the segmentations. 3. Confirm by clicking **Apply**. - **New audience** - Create audience by defining the conditions profiles must meet to be recipients of selected schema. 4. In the **Audience** section, click **Apply**. ### Reviewing schema settings In the **Audience & Settings** tab, you can preview the **API name** and schema description you defined at the beginning of the process. You can only edit the description at this stage. ### Saving schema In the upper-right corner, click **Save**.
The final configuration of a schema
The final configuration of a schema
## Next steps --- As the next step, you will [create a record](/docs/assets/brickworks/quick-start/creating-a-record) out of the created schema. ### Editing schemas When editing existing schemas, keep the following points in mind: - Modifying schema does NOT affect existing records. If the field has a default value, that value will be applied when content is generated from the record. - Removing a field from a schema does NOT remove it from existing records; but when content is generated from such a record, that value is ignored. - You can't open and view unpublished records created from a previous version of a schema.