Sweep AI

What is Sweep AI?

Sweep AI is an AI coding assistant that manages GitHub tickets and automates coding tasks. It turns bug reports and feature requests into pull requests, generating code changes across your entire repository. It integrates with Slack and lets you take action directly within GitHub. Teams such as PyTorch-Ignite and Medplum, along with universities including Stanford, Carnegie Mellon, and Waterloo, have used Sweep to maintain code quality, handle repetitive tasks, and get code suggestions.

Who created Sweep AI?

Sweep AI comes from Sweep.dev and was founded by William Zeng and Kevin Lu. William studied Stats & ML at CMU, and Kevin left UWaterloo CS; both previously worked at Roblox. They built Sweep to automate software development tasks with AI.

What is Sweep AI used for?

• Turns GitHub issues directly into pull requests
• Generates code changes across your entire repository
• Suggests fixes for your code
• Connects with Slack so you can preview its plans, request test runs, and get pull request alerts
• Supports code review and changes within GitHub
• Automates repetitive tasks and helps maintain code quality over time

Who is Sweep AI for?

Sweep AI suits software developers of all kinds: teams that use GitHub, projects focused on collaborative development, and independent developers looking to improve productivity and streamline their workflows.

How Sweep AI works

Connect Sweep to your repository, then file issues as you normally would. Sweep analyzes them and produces pull requests. Through Slack, you can review what it plans to do, request test runs, and receive alerts about its pull requests, keeping the workflow inside tools your team already uses.