# Retrieve job results
<a name="retrieving-results"></a>

After a job reaches **Completed** status, download the result files from your Amazon S3 output bucket. The service writes results in Apache Parquet format, preserving all original input data and adding results from the processed action.

## Output schema overview
<a name="output-schema-overview"></a>

Output files contain the following general field categories, regardless of the job action type:

Input fields  
All original input fields are preserved with an `Input_` prefix. For example, if your input included an `Id` field, it appears as `Input_Id` in the output. This allows you to correlate output records with their corresponding inputs.

Output\$1Id  
A unique identifier generated by the service for each output record.

Error fields  
If an error occurs during the processing of a record, the output includes `ErrorType` and `ErrorMessage` fields with details about the error. Records that process successfully do not include error values.

Action-specific output fields  
Each job action type adds its own result fields to the output. These fields contain the processed results for each record.

For address validation output schema details, see [Output schema](address-validation-concepts.md#address-validation-output-schema).

## Retrieve results from Amazon S3
<a name="retrieve-results-procedure"></a>

**To retrieve job results**

1. Open the Amazon S3 console at [https://console.aws.amazon.com/s3/](https://console.aws.amazon.com/s3/).

1. Navigate to your output bucket.

1. Locate the output files created by your job.

1. Select the files and choose **Download**.

Process the results according to your application needs.