For AI agents: a documentation index is available at the root level at /llms.txt and /llms-full.txt. Append /llms.txt to any URL for a page-level index, or .md for the markdown version of any page.
Sign inBook a demo
DocsReferenceChangelog
DocsReferenceChangelog
  • Getting Started
    • Overview
    • Quickstart
  • Explanation
    • Integrating Humanloop
  • Tutorials
    • Evaluate an Agent in the UI
    • Evaluate an Agent in code
    • Evaluate a RAG app
    • Capture user feedback
  • How-To Guides
    • Migrating from Humanloop
  • Reference
    • Deployment Options
    • Supported Models
    • Template Library
    • Vercel AI SDK
    • .prompt and .agent Files
    • Humanloop Runtime Environment
    • Security and Compliance
    • Data Management
    • Access roles (RBACs)
    • SSO and Authentication
    • LLMs.txt
LogoLogo
Sign inBook a demo
On this page
  • Data Handling and Segregation
  • Data Classification & Access Control
  • Data Encryption and Security
  • Encryption
  • Infrastructure
  • Learn More
  • Data Storage, Retention, and Recovery
  • Data Breach Response
  • Data Portability and Return
Reference

Data Management

An overview of the data management practices and encryption methodologies used by Humanloop
Was this page helpful?
Previous

Access roles (RBACs)

Next
Built with

Data Handling and Segregation

Separate environments are provisioned and maintained for development, quality assurance/user acceptance testing, and production to ensure data segregation at the environment level.

Data Classification & Access Control

All platform data received from the user and data derived from user data is classified as sensitive. All platform audit and telemetry data that does not contain PII and reference to specific user data is classified as not sensitive.

By default, only authenticated users can see their own sensitive data. Data classified as not sensitive can be accessed by dedicated Humanloop support staff using a secure VPN connection to the private network of the VPC for the target environment. This access is for debugging issues and improving system performance. The Terms of Service define further details around data ownership and access on a case-by-case basis.

Data Encryption and Security

Encryption

Humanloop follows best practices for data management and encryption. All data in transit is secured with TLS/SSL, and all data at rest is encrypted using the AES-256 algorithm. All encryption keys are managed using AWS Key Management Service (KMS) as part of the VPC definition.

Infrastructure

All sensitive data is encrypted in transit. For Self-Hosted Cloud (VPC) environments, network traffic is also encrypted in transit and at rest to meet HIPAA requirements. Sensitive application data is only processed within the ECS cluster and stored in Aurora. To request a network infrastructure diagram or more information, please contact privacy@humanloop.com.

Learn More

For more information on how Humanloop processes user data, visit our Security & Compliance page.

Data Storage, Retention, and Recovery

All platform data is stored in a primary database server with multi-availability zone replication. Platform data is retained indefinitely and backed up daily in a secure and encrypted manner until a request is made by the contractual owners of that data to remove it, in accordance with GDPR guidelines.

Humanloop’s Terms of Service define the contractual owner of the user data and data derived from the user data. A semi-automated disaster recovery process is in place to restore the database to a specified point-in-time backup as required.

Data Breach Response

Any data breaches will be communicated to all impacted Humanloop users and partners within 24 hours, along with consequences and mitigations. Breaches will be dealt with in accordance with the Humanloop data breach response policy, which is tested annually.

Data Portability and Return

Within 30 days post-contract termination, users can request the return of their data and derived data (as defined by the Terms of Service). Humanloop provides this data via downloadable files in comma-separated value (.csv) or .json formats.