
PopTask - AI Powered Menu Bar To-do List
Type a messy thought, get a structured task in 3 seconds
am haider, solo dev in jeddah .. this is the honest version
i built PopTask because every to-do app made me fill out a form for a 2-second thought .. title, date, time, repeat, reminder .. so i'd lose the thought or not bother .. i wanted the opposite: type the mess the way you'd say it ("gym mon wed fri 6am") and get a real scheduled task in about 3 seconds, no forms
the hard part was the parser .. a cascade that runs mostly on-device (apple's foundation models), handling typos, shorthand, multiple weekdays, in 9 languages, without a server round-trip .. that took most of the year
then reality: it's been live on mac and getting about 1 download a day for months .. no reviews, no backlinks, no press .. a brand-new app in the most saturated category there is .. that's not a keyword problem, it's an authority problem
so am doing the thing i kept putting off: a real coordinated launch, and i brought it to iphone + ipad too (one universal purchase, on-device parsing, own-icloud sync, no tracking) .. i built it for myself first, people made it stick, now it's my full-time focus
if you've launched a solo productivity app into a crowded category .. what actually moved the needle for you? and if you want to kick the tires, throw your messiest input at it and tell me where it breaks: https://apps.apple.com/app/id6759860524
About
every to-do app makes you fill out a form for a quick thought, so you lose it .. PopTask kills that: type it messy, get a real scheduled task in ~3s .. i built it for myself, now it's my full-time focus

1 Comment
What stood out to me is that the core problem wasn’t task creation—it was friction at the moment of intent. Most to-do apps assume users are willing to translate intent into structured fields, but real behavior is messy and time-sensitive. If PopTask consistently captures that “raw intent → structured action” conversion in under a few seconds, that’s not just a UX improvement, it’s a different interaction model entirely.