Multi-passionate creator. Pour Recall π· wine journal app β’ Wear Your Words π tees β’ IFIO Girl Learning Co π TPT leveled activities for every learner
61, no coding background, and I learned Apple and Google have new rules for AI in apps. My new AI Label Scan, powered by @OpenAI , asks before sending your photo, and a switch lets you change your mind anytime. #BuildInPublic#NoCode#AI
@vinos25@AdaloHQ@OpenAI@claudeai Thanks! Small correction: I wired in OpenAI's vision model myself, and Claude wrote the JSON. The hardest part was typing every { [ ] } : " by hand with a jeweler's precision. π I make many typos in everything I do! So, TestFlight was definitely the easy part!
61, no coding background, and today AI read a wine label inside my own app. π· @AdaloHQ built-in @OpenAI couldn't see images, so I built my own connection with @claudeai assistance. It passed the TestFlight test! Next up: App Store and Google Play. #BuildInPublic#NoCode#AI
@ndmrau@AdaloHQ@OpenAI@claudeai Thank you! Great tip. I did not even think about putting that in this submission. I will write a clear reviewer notes that can walk them straight to the new AI scan and the consent screen. π
@pushpendraw@Joshua_WD@AdaloHQ@OpenAI@claudeai Great tip, and good news, Pour Recall already saves the label photo with every pour. So when Region comes later, a "rescan" from the saved photo seems totally doable with a "rescan" button??? Thanks for the idea.
@deeponailabs@AdaloHQ@OpenAI@claudeai WOW! Thanks for this! Photo permissions are set, and it got me double-checking my AI disclosure too. It looks like Apple now wants explicit consent before sending data to third-party AI, so I'm adding that before submitting. many thanks!
@iamvishal16_ios@AdaloHQ@OpenAI@claudeai Honestly, the response format. Adalo can't split one AI answer into separate fields, so I use two calls: name and type. 2nd was typing in the JSON code @claudeai wrote for me. I make typos all the time, so I typed it with the precision of a jeweler. ππ·
@SamVale29@AdaloHQ@OpenAI@claudeai Totally agree, real-world testing beats any benchmark! Glare and curved labels are next. Wine Wednesday at @Cascones in Overland Park sounds like the perfect test lab.π· And if it misses, the name stays editable and the type can be changed, so it's a quick fix.
@topphamilia@AdaloHQ@OpenAI@claudeai That's exactly why I built it! Save the bottles you loved, and the ones you want to try next. Pour Recall is for people who just want to remember what they liked.
@Joshua_WD@AdaloHQ@OpenAI@claudeai Thank you! Right now it reads the wine name and type. Region and varietal mean extra AI calls, unless I route it through Make to do it all in one. That's a bigger build than I've tackled yet, but not out of reach. For now, starting lean and keeping costs low.
@BestAppAwards@AdaloHQ@OpenAI@claudeai Great point! Demo shots are the easy part. I'll put it through a real restaurant test next. The good news is the name stays editable and the type can be overridden, so a miss is a quick fix, not a dead end.
@AdaloHQ I see what other people have made using Adalo, and what I've done barely scratches the surface of what's possible. Still, my focus is keeping Pour Recall simple. But on my wishlist wall: "Hey Alexa, Hey Google, Hey Siri, log this wine into My Pour Choices." βΊοΈ
It worked. Right inside my real app, not a test screen. π· I choose a wine label photo, tap one button, and AI reads it, fills in the name, and picks the wine type, automatically. Built with no code: @AdaloHQ + @OpenAI + @claudeai#BuildInPublic#NoCode
@AdaloHQ Thank you, @AdaloHQ , for the continued support! Building Pour Recall (an IFIO Girl app) with no coding background hasn't been easy, but your team has helped me push through every wall. π· Keep wine sipping simple