# External trigger
The task "**External trigger**" needs to be placed at the very beginning of the pipeline if you want to execute it by calling [Ylem's API](https://docs.ylem.co/api/api-endpoints#run-pipeline) from the outside.
The following cases are the most common for such a situation:
* Trigger pipeline from [Apache Kafka](https://docs.ylem.co/integrations/library-of-integrations/apache-kafka), [AWS Lambda](https://docs.ylem.co/integrations/library-of-integrations/aws-lambda), [AWS S3](https://docs.ylem.co/integrations/library-of-integrations/aws-s3), [Google Pub/Sub](https://docs.ylem.co/integrations/library-of-integrations/google-pub-sub), [RabbitMQ](https://docs.ylem.co/integrations/library-of-integrations/rabbitmq)
* Trigger pipeline by [calling an API](https://docs.ylem.co/api/api-endpoints#run-pipeline) directly or calling it from your custom script
The input dataset that you transfer to the API endpoint will also be available in the task that follows "**External trigger**":
The following simple pipeline example accepts data from outside and forwards it further as a notification.
Let's send the following JSON as input to the [API endpoint](https://docs.ylem.co/api/api-endpoints#run-pipeline) that triggers this pipeline:
```json
{
"info": "some information",
"customer_id": 1
}
```
As you can see the input dataset is available in the "Notification task" as a whole under the `{{ INPUT() }}` function and also every input key-value pair is available separately by its key name.
