Powering the next generation of AI software agents

We're excited to announce our partnership with Tusk to bring AI testing superpowers to their AI coding agent and launch Automated Browser Testing.

Tusk is an AI coding agent that writes, pushes, and tests code for front-end tasks. Engineers use Tusk to automatically fix customer-reported bugs and make UI/UX improvements, increasing customer NPS while reducing context-switching.

The future of AI-generated software

When we pitched Momentic to hundreds of investors, One of the core pillars of our vision was how testing was going to play a major role in the future of software. In a world where more and more code is generated by tools like Cursor, GitHub Copilot, and others, it's become apparent that the quality simply isn't there.

Tusk is the first AI coding agent that can independently verify its changes in a live browser. The Tusk team learned it was time-consuming for engineers to verify whether generated PRs could be merged. They needed a way to make the PR review process easier while enabling their agent to generate higher-quality code. They did this by creating a tight feedback loop between testing and code generation.

How it works

After generating a PR for a task, Tusk will generate the English instructions required to verify the expected behavior. Tusk then leverages Momentic's AI testing platform to execute the browser test on your PR's preview environment.

By doing so, their coding agent emulates a human engineer using a live browser, iterating based on the results, and leaving screenshots/videos as proof of completion in the PR to make reviews easy.

To leverage Automated Browser Testing today, record a video showing how to get to the location of the UI change and/or how to replicate a bug. You can also narrate over the video such that you don't need to write a detailed issue description.

For more details, check out their documentation or reach out to founders@usetusk.ai.

Web agent and browser infrastructure

Automated Browser Testing builds on top of Momentic's proprietary web agent. You can provide instructions in natural language and replicate user flows live in a browser on-demand.

Our infrastructure executes hundreds of thousands of web and API tests each month with 99.99% uptime and reliability. We're prioritizing access to our API to companies pushing the boundaries of AI in software.

If you'd like to leverage Momentic's AI testing platform, reach out to us at sales@momentic.ai.