Browser automation is needed in many situations. Though there are numerous ways to get it rolling: self-hosted or cloud and ai or selectors-based software. All those approaches can be handy in some cases, while pretty useless in others. So one has to understand what's the use case for them.
What I needed is a straightforward API/webhooks point to send tasks to predefined AI browser agent tasks. I have some selector-based zenno scenarios and they work great! But still, as soon as the website changes they have to be updated and sometimes that results in hours of the functionality down.
My tests with Browseract are fully satisfying. The product lacks some extended features like hard-coded loops and persistent browser instances. Still, those are on their roadmap and the team has been very helpful and fast in their support. As with the majority of the AP deals, you can get a very promising software for a fraction of the market cost. You will have to wait for some needed features to be released,(which rather means, you just adapt your tasks so they're doable in the current software state so they're not fully optimized but still work) but in the end you are rewarded.
Thank you