We're scaling our early sales and seeking QA engineers, customer support, and sales engineers.

Please DM ian [at] testdriver [dot] ai

The problem is not the docs, it's Conway's law. One team designs the API, the other team designs the portal, and another team designs the SDK. The user has a holistic experience that cuts through each team.

That, and the docs are usually written first by the most technical person around, who has a hard time sharing the world view of a noob.

This is the late game, why would an engineer work for a fraction of a percent of equity and a below market salary when they can take a job at FANG?

You've got to be offering something really, really valuable like remote work, an interesting problem, and/or a new experience. Otherwise the math doesn't math.

We've built an AI Agent that performs manual testing on it's own VM with complete desktop access. It works like a specialized "Claude Computer Use."

We're scaling our early sales and seeking QA engineers, customer support, and sales engineers.

Please DM ian [at] testdriver [dot] ai

We have multiple fallbacks to prevent flakes; The "cheap" command, a description of the intended step, and the original prompt.

If any step fails, we fall back to the next source.

1. https://docs.testdriver.ai/reference/test-steps

95% of companies are still wasting time manually testing due to shortcomings in Playwright, Cypress, and other frameworks. Developers rank testing as the #1 blocker to release.

We've built an AI Agent that performs manual testing on it's own VM with complete desktop access. It works like a specialized "Claude Computer Use."

We're scaling our early sales and seeking QA engineers, customer support, and sales engineers.

Please DM ian [at] testdriver [dot] ai

Copy and demo videos are essentially one way communication channels ("fire and forget"). The creator has no idea if the message was understood.

Also, writing copy or making a video typically takes 10 - 100x longer than consuming the same video.

An average typing speed is 40wpm but an average conversation is between 120 - 150 wpm so about 3 - 4x bandwidth.

Calls also offer sub second latency and maximum priority.

When you add video and audio in there, the pure amount of data transferred is higher.

95% of companies are still wasting time manually testing due to shortcomings in Playwright, Cypress, and other frameworks. Developers rank testing as the #1 blocker to release.

We've built an AI Agent that performs manual testing on it's own VM with complete desktop access. It works like a specialized "Claude Computer Use."

We're scaling our early sales and seeking QA engineers, customer support, and sales engineers.

Please DM ian [at] testdriver [dot] ai

Comments URL: https://news.ycombinator.com/item?id=42392209

Points: 1

# Comments: 1

- A list of applications that are open - Which application has active focus - What is focused inside the application - Function calls to specifically navigate those applications, as many as possible

We’ve found the same thing while building the client for testdriver.ai. This info is in every request.

Comments URL: https://news.ycombinator.com/item?id=41927314

Points: 9

# Comments: 1

Yes, you are correct that it entirely lays in the reputation of the AI.

This discussion leads to interesting question, which is "what is quality?"

Quality is determined by perception. If we can agree that an AI is acting like a user and it can use your website, we can assume that a user can use your website and therefor it is "quality".

For more, read "Zen and the Art of Motorcycle Maintenance"

Another example, imagine an error box shows up. Was that correct or incorrect?

So you need to build a "meta" layer, which includes UI, to start marking up the video and end up in the same state.

Our approach has been to let the AI explore the app and come up with ideas. Less interaction from the user.