HumanLayer is an API and SDK that enables AI Agents to contact humans for help, feedback, and approvals. It’s perfect for:

  • Implementing manual approval steps for software- and human-driven workflows
  • Oversight of autonomous AI agents
  • Managing the transition between the two

​
Getting Started

Get started with Humanlayer using Python or Typescript.

Python Quickstart

Get started with Humanlayer using Python

Typescript Quickstart

Get started with Humanlayer using Typescript (COMING SOON)

​
Features

Core Features

Explore the core features of HumanLayer like approvals, human-as-tool, and more

Contact Channels

Learn about the different contact channels that HumanLayer supports like Slack, Email, React Embeds, and more

Frameworks

Learn about the different frameworks that HumanLayer supports like OpenAI, Langchain, CrewAI, and more

API Reference

Dig into the details

​
Introduction to HumanLayer

​
Key Features

  • Human Approvals: Integrate human oversight into critical AI decisions
  • Flexible Workflows: Design custom approval workflows with multiple reviewers
  • Timeout Management: Set appropriate timeouts for human responses
  • Integration Ready: Easy integration with existing AI applications
  • CLI Tools: Command-line interface for common tasks

​
Why HumanLayer?

As AI systems become more prevalent, the need for human oversight and quality control becomes increasingly important. HumanLayer provides the infrastructure needed to:

  • Ensure quality control in AI-generated content
  • Maintain human supervision over critical AI decisions
  • Create auditable approval workflows
  • Integrate human expertise with AI capabilities
  • Collect and export human feedback into evaluation or fine-tuning datasets

​
Getting Started

Check out our quickstart guides:

​
Installation

pip install humanlayer
Python Quickstart