Corkdork

Wine tasting notes & journal

2 followers

Master wine tasting with Corkdork by taking guided, step-by-step wine tasting notes while drinking a glass of wine—no experience or training necessary. Track the wines you taste over time and understand why you like what you like with analytics. 21+.
Corkdork gallery image
Corkdork gallery image
Corkdork gallery image
Corkdork gallery image
Corkdork gallery image
Corkdork gallery image
Corkdork gallery image
Launch tags:iOSiPadWine
Launch Team
AssemblyAI
AssemblyAI
Build voice AI apps with a single API
Promoted

What do you think? …

Clay Heins
@rueter This is very interesting, I have been trying to find something like this for tastings and notes. Have you looked more into recognition of the text on the labels to autofill some of the inputs?
Steven Rueter
@clay_heins hi Clay! Thanks for your comment. I have looked into a few label/image/text recognition APIs. There is currently a combination of two working in the background: a text recognition ML that can pick up on and instantly autofill legible text, and a more nascent image recognition ML model being trained in the background to eventually be able to specifically pick up on certain labels. If you have any good APIs you think are worth looking at, please let me know.
Clay Heins
@rueter One thought is using Google's Cloud Vision API: https://cloud.google.com/vision/... . I am not sure this would work with your app, but might be worth taking a look at. There are more custom solutions like OpenCV and Tensorflow that could be good at label detection and text recognition. If you have something that is already autofilling text then maybe these won't add much!
Steven Rueter
@clay_heins thanks for getting back to me. The app does already utilize Cloud Vision to a degree. The big thing is getting enough labels to train to model to recognize them. Wine labels are a particularly interesting sample set for ML image recognition because, as many as there are out there, it's still just a finite set. My hope is to make a UX good enough that will encourage users to add labels manually until the model is trained enough to take over 🦾
Steven Rueter
Corkdork has been my nights-and-weekends project since the shelter-in-place/WFH orders took effect in March. Corkdork teaches you the wine tasting process, tracks the wines you taste, and in the long run, helps you understand why you like what you like all with the goal of developing your own tastes and preferences. I'm glad to finally be able to share it with the PH community and look forward to your feedback.
Steven Rueter
@nazmul_husain thanks glad you like it. Feel free to reach out to @corkdorkorg on Twitter and Instagram.
Oliver Boyers
This is cool! Do you have image recognition of labels? I was curious about the tech used to facilitate that!
Steven Rueter
@oliver_boyers thanks for your comment. The app is using a combination of Google’s ML Kit SDKs. For these early days, however, it relies mostly on user entry for label information, which poses a UX challenge. To address that, I’ve primarily focused on creating a seamless stepwise form entry UI/UX. This works, though, since the rest of the tasting process, namely the Look, Smell, and Taste sections require the same style of active user input.
Katheriin Liibert
Interesting! Been a big fan of Vivino for the past 8+ years (🥂), so curious to check this one out.
Steven Rueter
@niirehtak Thank you, I'm looking forward to hearing what you think ☺️🥂
Dante Ventu
When is it coming to android :(
Steven Rueter
@danteventu thanks for your comment. We'll see if there is enough interest for the product and then will happily port to Android 🍷
Charlotte Franssen
This is cool! Love the name as well! :)
Steven Rueter
@charlotte_franssen thank you 😊
Jamie Wells
💯% recommended
Steven Rueter
@jamiewells4 thanks, Jamie 🙏
12
Next
Last