I previously built a digital life calendar that helps me visualize how I spend my time, track my goals, and remind me about them at the right time.
Next, I wanted a journal app to complete my productivity stack. I started with ChatGPT to build an HTML tool that I could run in a browser.
But it required an API key for the AI features to work, and that’s how I ended up in Google AI Studio.
I got what I wanted, but there was something much more interesting. The platform now allows you to build full-fledged Android apps with simple prompts.
I asked AI to build the app, and it actually did
Google AI Studio surprised me with a working Android app
I briefly described the features I want in the text box and asked it to “add features that make sense.” It took about 10 minutes to think through and build the first version.
Finally, I had to get to my desk. Up until this point, I had been doing everything from my phone.
For the first time, I used the USB debugging feature in Android’s Developer Options to install the app Gemini had built directly via a USB cable.
Unlike Claude, which often runs out of tokens before finishing complicated projects, Google AI Studio finished and got everything working as a fully functional Android app.
I was surprised to see that it introduced useful ideas I hadn’t specifically asked for and had never seen in a journal app before.
The interface, however, had room for improvement.
The first version had more than one button that did the same thing. In many cases, it used plain text as a touch target instead of a button-like design. There were also some features I did not want.
So I followed up with AI Studio, asking for a Material You-style look, Google’s proprietary design language, and the removal of the few features I did not want.
After three hours of fine-tuning, I had something that convinced my friend it was from the Play Store.
I previously tried installing HTML tools as apps on my phone. I could only package them as lightweight APK wrappers, but they still ran in the browser, which limited what the tools could do.
This time, I got a full-fledged Android app from a single prompt, which was surprising.
I didn’t want another blank text box
AI turned journaling into something I can actually revisit
Let’s talk about this tool’s features. At its core, it offers a text box where I can write an entry every day. But it’s easy to get sidetracked when staring at an empty text box.
Remember how I asked it to add features? It seems to have taken that request seriously. When I add an entry, the page is packed with useful features.
First, there are separate mood options, such as grateful, energetic, accomplished, and reflective. There’s also an energy index, where I can rate my energy from 1 to 5.
Then there are three separate text boxes. The first is “What happened?” or what I did that day.
The second is “Deep Reflection & Feelings,” so I can separate the facts from my emotions. That’s helpful when I want to look back at my entries later.
And lastly, there’s a short section called “Daily Win,” where I can write a single line explaining the best thing about my day.
At the end of the page, Google AI also added tags and categories, making it easier to find entries later.
I can tag things such as work, focus, and habits. Tapping the All Entries button on the journal page and typing a tag brings up all the associated entries.
There’s also an option to add a custom tag, which makes it even more helpful.
I actually asked AI to make it easy to search through my entries, and I like how it implemented this feature.
One tap gives Gemini access to my entire journal
The convenience is also where the privacy question gets interesting
I haven’t started on the AI features yet. It’s easy to get overwhelmed when I want to look back at my entries. But we have AI, right?
Thanks to the integrated Gemini API key, this app, called “Reflect,” can give Gemini access to the entries so it can analyze them and find patterns I’d probably never take the time to spot myself.
I only need to tap the Re-Scan Patterns button on the Find Patterns page. As a cherry on top, I can also ask follow-up questions related to the patterns.
Technically, I could access these AI features through a chatbot, but I would need to share all the information with it beforehand. The context is important here.
Thanks to the API integration, it has everything it needs. I just need to allow access when I want to use the feature.
But that convenience raises an uncomfortable privacy question: Would I normally hand an AI my entire journal?
I would not enter any information in this app that I wouldn’t be comfortable sharing with a chatbot interface.
The more personal context you give an AI, the more useful it can become. But the privacy trade-off becomes bigger too.
At the end of the day, I feel like it’s about finding the right balance in how much to share, rather than keeping or removing the API integration altogether.
Why build this when journal apps already exist?
AI let me make the app I wanted instead of paying for someone else’s
There are already apps available on the Play Store with similar API-integrated features. They could be a good option, but most come with fairly expensive monthly or yearly subscriptions.
I wanted to check out the AI features, but none of the apps I tried allowed me to do so without paying for a subscription.
After that experience, I wanted to see if I could build my own free app that also integrates AI. As it turned out, I could use a Gemini API key to make it happen.
It was just a fun project I started late in the evening. I had a working app, convincing enough that it looked like something from the Play Store, before I called it a day.
What’s more interesting is that it has the exact design elements and feature set I wanted. If a new feature pops up on my wish list tomorrow, I can get it done within a few minutes just by asking.
I can install the app on any Android device I have physical access to, and I can even set up a Play Console developer account by following a few easy steps to see if it gets approved for the Google Play Store.
