# How to create a pipeline
Pipelines are created within the Canvas environment. Use the following steps to learn how to create, configure, and build your pipeline.
## **Creating the pipeline**
Follow these steps to create a pipeline:
1. On the Build page, click the **Pipeline** tab.
2. Click **Create new** in the upper right corner.
3. Select the **Pipeline** option.
The pipeline is automatically created within your **default project**. To create a pipeline in a specific project, select the desired project before clicking **Create**.
## **Configuring the pipeline**
1. Click the gear icon in the upper-right corner of the screen (next to the **Save** button) to access the pipeline settings.
2. A configuration panel will open on the right side of the screen. Fill in the fields as described below:
* **Description:** Enter a brief description of the pipeline.
* **Is it multi-instance?:** Enable this option if the pipeline should run as a multi-instance pipeline. For more information, see the [Multi-Instance](https://docs.digibee.com/documentation/developer-guide/platform-administration/settings/multi-instance) documentation.
* **Sensitive field:** Define the data fields that should be obfuscated in the pipeline logs, which will appear as `***`. The hyphen character (`-`) is allowed in the field name, but other special characters, accents, and the letter *ç* are not supported.
{% hint style="info" %}
Sensitive fields defined in this configuration apply only to this specific pipeline. If you need to apply sensitive fields to all pipelines in your realm, refer to the [Sensitive Fields Policy](https://docs.digibee.com/documentation/developer-guide/platform-administration/governance/policies/security/sensitive-fields).
{% endhint %}
* **InSpec & OutSpec (optional):** Use these configurations to define structural validation rules for the pipeline’s input and output in JSON format.
#### **InSpec & OutSpec: Validating the pipeline structure**
The **InSpec** and **OutSpec** settings allow you to validate the structure of incoming and outgoing data in **JSON format**. They help ensure that your pipeline only processes payloads that match the expected schema.
* **InSpec:** Validates the pipeline input payload structure.
* **OutSpec:** Validates the final output payload structure.
Both use JSON validators (similar to JSON Schema) to define the expected data format.
**Benefits**
* Prevents execution errors caused by unexpected data structures.
* Reduces rework due to system incompatibilities.
* Improves data flow security and predictability.
* Simplifies debugging via automatic schema validation.
* Documents input/output contracts clearly.
**How it works**
* Both **InSpec** and **OutSpec** contain a **JSON Validator** schema.
* If the payload violates the defined structure during execution, the pipeline is **interrupted**.
**Example: InSpec**
```json
{
"type": "object",
"required": ["id", "name", "email"],
"properties": {
"id": { "type": "string" },
"name": { "type": "string" },
"email": { "type": "string", "format": "email" }
}
}
```
This example ensures that the input payload includes `id`, `name`, and `email`, with a valid email format.
**Important notes**
* These fields are **optional**. Pipelines work normally without InSpec and OutSpec.
* They are recommended for high-criticality pipelines, multi-consumer flows, B2B integrations, and internal contracts between teams.
* They help improve integration reliability and data consistency.
## **Building the flow in the pipeline**
Follow these steps to create a flow in a pipeline:
1. Move the mouse pointer over the trigger already available in the Canvas and click on the **Configuration** button (represented by a gear icon).
2. Select a [trigger](https://app.gitbook.com/s/EKM2LD3uNAckQgy1OUyZ/triggers/overview) from the available options. The trigger starts the execution of the pipeline, such as an [event](https://app.gitbook.com/s/SKBJ6ZiEWBU93x170HH4/triggers/messaging-and-events/event) or an [email](https://app.gitbook.com/s/SKBJ6ZiEWBU93x170HH4/triggers/web-protocols/email).
3. Configure the selected trigger and click on **Confirm** to save your changes.
4. After configuring the trigger, click on the **+** button to begin creating your flow. Refer to the [Smart Connector User Experience documentation](https://docs.digibee.com/documentation/developer-guide/development-cycle/build-overview/canvas/smart-connector-user-experience) for more details about available connection options.
5. While building your flow, use the [**Execution panel**](https://docs.digibee.com/documentation/developer-guide/development-cycle/build-overview/canvas/execution-panel) for testing and the [**Inspect mode**](https://docs.digibee.com/documentation/developer-guide/development-cycle/build-overview/canvas/design-and-inspect-mode) for troubleshooting.
{% hint style="warning" %}
Pipelines are not saved automatically. You need to click the **Save** button each time you want to save your changes. The first time you save, provide the **Name**, **Description**, and **Project** to which the pipeline belongs.
{% endhint %}
The Canvas, the environment where pipelines are built, offers several navigation options. To learn more, read the [Canvas documentation](https://docs.digibee.com/documentation/developer-guide/development-cycle/build-overview/canvas).
### Pipeline size restriction
Each pipeline has a maximum size limit of **1 MB**, which includes all its components: [connectors](https://app.gitbook.com/s/EKM2LD3uNAckQgy1OUyZ/connectors/overview), [triggers](https://app.gitbook.com/s/EKM2LD3uNAckQgy1OUyZ/triggers/overview), [accounts](https://docs.digibee.com/documentation/developer-guide/platform-administration/settings/accounts), [globals](https://docs.digibee.com/documentation/developer-guide/platform-administration/settings/globals), and more. If the total size exceeds this limit, deployment will not be possible. Before deploying it, ensure that the pipeline size is within the allowed limit.
#### How to identify the size of your pipeline?
**Step 1:** Copy all the pipeline connectors using one of the following methods:
* Press **Ctrl/Cmd + A** and then **Ctrl/Cmd + C**.
Or
* Hold **Ctrl/Cmd**, click on each connector, and then press **Ctrl/Cmd + C**.
**Step 2:** Paste the connectors on [JSON Size Analyzer](https://www.debugbear.com/json-size-analyzer) pressing **Ctrl/Cmd + V** and execute a query to check the approximate size of the pipeline.
If the pipeline is smaller than 1 MB, it can be deployed. If it exceeds this limit, reduce its size before deployment.
