---
title: Add prediction data post-deployment
description: How to add historical prediction data after a model is deployed.

---

# Add prediction data post-deployment {: #add-prediction-data-post-deployment }

Users with the [Owner](roles-permissions) role can add historical prediction data to external model deployments if data drift is enabled. To do so, navigate to the [**Data Drift > Settings**](data-drift-settings) tab and click **Choose file** in the **Inference data** section to upload your prediction data in XLSX, CSV, or TXT format. You can also select prediction data from the **AI Catalog**.

 ![](images/add-historical-pred-data.png)

Training data is a critical component for calculating data drift. If you did not include training data when you created a deployment, or if there was an error when uploading that data, you can also add training data from the [**Data Drift > Settings**](data-drift-settings) tab.

These datasets must meet the following requirements:

* **Historical prediction data**: The uploaded data have the same features as the original prediction dataset. After uploading new data, DataRobot prompts you to confirm the addition because you cannot remove data from a deployment later. To use different prediction data, create a new deployment.

* **Missing training data**: The uploaded data must include the same features as the prediction (scoring) dataset. You cannot replace training data. If you want a deployment to use different training data, create a new deployment with the appropriate data.
