---
title: "What is Coolhand's 90-day data retention policy?"
answer: "After 90 days, Coolhand automatically removes the content of your LLM logs and feedback, including prompts, responses, and explanations. De-identified records and metadata, such as token counts, latency, and scores, are retained. This process is irreversible and applies to all customers."
category: "Logs \u0026 API"
tags: []
published: 2026-10-23T17:28:49Z
updated: 2026-10-23T17:27:38Z
canonical_url: https://coolhandlabs.com/help/what-is-coolhands-90-day-data-retention-policy
index_url: https://coolhandlabs.com/help.md
---

# What is Coolhand's 90-day data retention policy?

After 90 days, Coolhand automatically removes the content of your LLM logs and feedback, including prompts, responses, and explanations. De-identified records and metadata, such as token counts, latency, and scores, are retained. This process is irreversible and applies to all customers.

## The 90-day content scrub

To protect your privacy and manage data storage, Coolhand has a 90-day data retention policy. After 90 days, the content of your [LLM logs](https://coolhandlabs.com/help/llm-request-log-fields.md) and [feedback](https://coolhandlabs.com/help/feedback-collection-best-practices.md) is permanently removed. This process, known as the "content scrub," is automatic and irreversible.

The scrub removes the following fields:

*   **Prompts and responses:** `user_prompt`, `system_prompt`, `assistant_response`
*   **Feedback:** `explanation`, `original_output`, `revised_output`
*   **Other content:** `error_details`, `thinking_response`

## What is kept

While the content is removed, Coolhand retains de-identified records and metadata. This allows you to continue analyzing trends and performance over time without storing sensitive data.

The following information is kept:

*   **Log metadata:** model, provider, token counts, latency, cost inputs
*   **Feedback metadata:** `like`/`dislike`, sentiment, and other scores
*   **Identifiers:** `creator_id`, `creator_unique_id`, `client_unique_id`, `coolhand_fingerprint_id`, and `llm_provider_unique_id`

**Important:** User-provided identifiers are **not** removed during the content scrub. If you use these fields to store personally identifiable information (PII), it will not be removed by this process. See [What every field on an LLM request log means](https://coolhandlabs.com/help/llm-request-log-fields.md) for a complete reference of all log fields.

## How it works

The content scrub runs automatically for all Coolhand customers and for all logs and feedback. Once the content is removed, it cannot be recovered. The scrub is designed to be a seamless process that requires no action on your part.


---

Related:
- [What every field on an LLM request log means](https://coolhandlabs.com/help/llm-request-log-fields.md)
- [LLMs and System One models: what's different, and when to use which](https://coolhandlabs.com/help/llms-vs-system-one-models.md)
- [How do I list and analyze my workloads using the API?](https://coolhandlabs.com/help/list-and-analyze-workloads-api.md)
- [Python SDK Integration Best Practices](https://coolhandlabs.com/help/python-sdk-integration-best-practices.md)

[← All help articles](https://coolhandlabs.com/help.md) · [Logs & API](https://coolhandlabs.com/help/category/logs-and-api.md)