Jots: AI summarized voice journal

[abstract hatch]

A iOS journalling app that uses Whisper to recognize input text and then uses a GPT-4-based model to summarize the text, generate a title and detect a mood.

Happy to answer questions and heear feedback 🙂

https://apps.apple.com/us/app/jots-ai-journal-diary/id6450964816

[noble zealot]

Please adopt this to Android

[abstract hatch]

It's a cross-platform app, so we're going to release this for Android as well 🦾

[odd moon]

Just tried it out, I like it! I would love to see an Apple Watch app at some point for easier note taking.

[abstract hatch]

Great idea! I always wanted to find a use case to develop something for Apple Watch and this could be it 🙂

[twilit mirage]

I just installed the app and congratulations on the app!!! The UI is user friendly and Jots is a really creative and potentially impactful use of the APIs.

I think you've got something great going on, here's some UX feedback:

App + Marketing Messaging
As technologists, we are very excited about technology but Jots' target audience could be less familiar with AI. I would suggest prioritizing the values people get from using Jots over the technology it was built with. Think about your target users, what does Jot uniquely provides the answers to their need (ex. Reflection, “Start your journey of self-reflection, growth, and understanding today with Jots: AI Journal & Diary. It's time to transform your thoughts into insightful summaries and unravel the depths of your mind.” <— this is the revolution that makes Jots better than any journal out there)

Privacy: That was on my mind when using the app since journaling is a very personal thing. I think I saw a brief message regarding the data being store on the phone. Personally I had questions on how and if my data would be share with 3rd party/OpenAI and will be use. Could be good to clarify that as privacy is a selling point (example: Apple).

Mood detection: Super interesting use of summarization. Have you consider letting people edit the emotion? Ex. Show the detected mood but the user can tab on it and a list of emoji shows up for the user to pick one. That could give user a sense of control if they don't agree with the sentiment and gives you data points on the mood detection accuracy. OpenAI could incorporate mood summarization into Whisper because tone of voice is revealing for mood detection than just text.

Ask AI: Consider having “prompt” on the Ask AI tab. On the App Store you wrote “Conversational Insights: Ask Jots about any of your past entries, and our AI will provide immediate, insightful feedback, enhancing your self-awareness journey.” Which is useful to guide and instruct your user on the intention of the tab. Considering placing a few prompt/starter message for user to kick off their conversation with AI. As a first time user, I didn’t know what to do in this section as it was blank with a text input area, so I asked “How to best use jots” and the answer made jots sounded like a note/task management app.

[twilit mirage]

Hi @abstract hatch the app works well and the UI is nice and polish. I’m wondering if you have any tips and lessons from working on this app and also what people should look for when looking to work with developers on building apps with the API (ex difficulty level, time it takes etc). Thanks and really appreciate it

[distant fjord]

@abstract hatch would love to talk about collaboration :)

[abstract hatch]

Didn't actually see your message. Feel free to contact me 🙂

[abstract hatch]

Thank you 🙂 Not sure what kind of advice is helpful to you, but I think it's quite key to go and try to get something out of the door as quickly as possible. We created the first prototype for this in a day and submitted it to the AppStore the next day and had it approved and live the next day. Then you can go and improve on it. Generally GPT/Whisper APIs are fairly simple and a free Vercel plan with edge functions is a good way to get started. Let me know if you have other questions.