A couple of weeks ago, I remember stressing over the launch with my cofounder because one random fine day I told him, "btw... I scheduled a launch for 23rd next week" and we were not at all prepared for what was about to come <insert hysterical laughter>
We had the month watching other launches and trying to figure out they did different to get a top launch, visiting product hunt every day and studying the launches, who the makers were, what their product did, their positioning, creatives and more.
Locking a job in once it runs right is the part I'd have led with. Re-planning a solved task on every run is where credits actually disappear, and most agent products quietly bill you twice for the same thinking. What I'd want to know is what happens when the locked path breaks because an API changed under it. Does he notice and re-plan, or run the stale steps and report success?
@asadmalik901 Hey Asad, valid question. When a workflow fails, the respective agent gets a failure notification, and it lets the workspace owner or the responsible user about it, they may then appropriately tell the agent to fix it.
Congrats on the launch! Giving AI agents dedicated workspace environments instead of clogging local systems is a smart approach to hands-off execution. Curious to see how it handles long-running, multi-step browser tasks over time.
@thisiskp_ Thanks KP! We built our own harness from first principles to run as natively as possible on cloudflare, and extensively use alarms+DOs+workers to support long running + recurring agentic tasks, as well as incorporate kitesurf + browser use for agentic browsing and web use as per requirement of the task.
btw a while back, we were even working on a desktop companion, which the cloud agent could handover tasks to and would even have the possibility of using desktop apps for things cloud agents may not be efficient enough, tho we are still experimenting with the UX side of things, the possibilities are endless.
WOuld love to hear your views on this 🙌
Congrats on the launch @ankushkun & team!
We run most of our GTM at Brew through agents and honestly a big chunk of the bill is the thing reasoning its way to the same conclusion it already reached last Tuesday.
Will give it a go this week!
@philip_sorensen Thanks Philip! Excited for you to give Construct a spin.
Would genuinely love to hear your feedback, and how it compares to your existing agent workflows :)