Its been around one year since I started thinking about creating apps or websites in my free time. However, as a master student in mechatronics, I was too addicted to the projects I was working on at university, and didn't dedicate time to programming as a side project. After one year of the master (~12h per day), I started working during summer 2025. This meant 8h per day of work and then I had the afternoons free. This was my moment to start building something, finally!
The first week of June, I was in a one-week course in Brussels (about control systems), and I also visited the city. While walking in the streets, I wondered about an app that could take a photo of a monument and get an audio answer, something like an AI tourist-guide. I was thinking that this was a pretty simple idea, and that probably somebody had allready done it. To my surprise, I checked it out on the internet, but didn't really find nothing similar, so I just added it to my huge list of app ideas.
Building time baby - First failure:
Back home from the trip, back to the routine, back to start building. I had another promising idea for a desktop app, so I started coding it with cursor. By the way, I knew programming (from the masters), but I had no idea about programming apps/websites -> with Grok and Cursor I was able to learn a lot. After almost one month of coding in the afternoons and in the weekends, I more or less had an MVP, but it was a complete failure, so I just decided to move on.
Keep grinding - Pivoted to Photo Tour AI:
In this way, in mid July, I started building Photo Tour AI with Flutter in Android Studio. This time I was mainly using Grok for coding assistance, but I found more helpful tutorial videos on Youtube, as well as following documentations. I was enjoying the code much more this time, as I was coding everything myself and understanding where the things were located. In the previous app using Cursor, I more or less knew where the thing where, but no exactly sure of what was going on, which in part made me fell less comfortable.
App done - Slow Deployemnt Period:
In the end of August, I had the app working, so it was time for deployment. This took me more time than I originally thought, as Google Play asked for 14 days of testing with 12 testers.
While waiting, I was trying to run macOS in a VM from Windows. After 5 days of struggling with the installation and trying different XCode/macOS versions, I achieved to set it up. Unfortunately, it was way too slow (it was taking one hour to just run the app in debug mode, and still did not run), so I decided to buy a macbook (hard decision, long term investment). From the macbook, I was able to develop and build for iOS, for deployment in the App Store. I also got a bunch of feedback from friends and family, which heIped me a LOT. I fixed a lot of bugs and even made some changes in some functionalities.
What it does now:
Now, after all the waiting, I'm finally able to share Photo Tour AI, available for Android and iOS.
The app lets the user take a picture of anything (monument, sculpture, building, mountain, tree, door, etc) and returns an AI generated audio, answering like a local tourist-guide to what it can see in the image. In addition, the user can make questions to Photo Tour AI, in case there is something specific that he wants to know.
The main target audience for the app right now are english speaking tourists. I will expand to more languages if it works in english :)
I need your help:
Solo founder, zero audience. Would love feedback from people who actually like to travel or love AI.
The app is free to try: 2 free images/questions per day.
Now I use the app in every city I go to, even in my hometown, where I'm discovering new things I didn't know about.
Feel free to ask any question, I will tell you everything I learned along the way.
jonurce•1h ago
Its been around one year since I started thinking about creating apps or websites in my free time. However, as a master student in mechatronics, I was too addicted to the projects I was working on at university, and didn't dedicate time to programming as a side project. After one year of the master (~12h per day), I started working during summer 2025. This meant 8h per day of work and then I had the afternoons free. This was my moment to start building something, finally!
The first week of June, I was in a one-week course in Brussels (about control systems), and I also visited the city. While walking in the streets, I wondered about an app that could take a photo of a monument and get an audio answer, something like an AI tourist-guide. I was thinking that this was a pretty simple idea, and that probably somebody had allready done it. To my surprise, I checked it out on the internet, but didn't really find nothing similar, so I just added it to my huge list of app ideas.
Building time baby - First failure:
Back home from the trip, back to the routine, back to start building. I had another promising idea for a desktop app, so I started coding it with cursor. By the way, I knew programming (from the masters), but I had no idea about programming apps/websites -> with Grok and Cursor I was able to learn a lot. After almost one month of coding in the afternoons and in the weekends, I more or less had an MVP, but it was a complete failure, so I just decided to move on.
Keep grinding - Pivoted to Photo Tour AI:
In this way, in mid July, I started building Photo Tour AI with Flutter in Android Studio. This time I was mainly using Grok for coding assistance, but I found more helpful tutorial videos on Youtube, as well as following documentations. I was enjoying the code much more this time, as I was coding everything myself and understanding where the things were located. In the previous app using Cursor, I more or less knew where the thing where, but no exactly sure of what was going on, which in part made me fell less comfortable.
App done - Slow Deployemnt Period:
In the end of August, I had the app working, so it was time for deployment. This took me more time than I originally thought, as Google Play asked for 14 days of testing with 12 testers.
While waiting, I was trying to run macOS in a VM from Windows. After 5 days of struggling with the installation and trying different XCode/macOS versions, I achieved to set it up. Unfortunately, it was way too slow (it was taking one hour to just run the app in debug mode, and still did not run), so I decided to buy a macbook (hard decision, long term investment). From the macbook, I was able to develop and build for iOS, for deployment in the App Store. I also got a bunch of feedback from friends and family, which heIped me a LOT. I fixed a lot of bugs and even made some changes in some functionalities.
What it does now:
Now, after all the waiting, I'm finally able to share Photo Tour AI, available for Android and iOS.
The app lets the user take a picture of anything (monument, sculpture, building, mountain, tree, door, etc) and returns an AI generated audio, answering like a local tourist-guide to what it can see in the image. In addition, the user can make questions to Photo Tour AI, in case there is something specific that he wants to know.
The main target audience for the app right now are english speaking tourists. I will expand to more languages if it works in english :)
I need your help:
Solo founder, zero audience. Would love feedback from people who actually like to travel or love AI.
The app is free to try: 2 free images/questions per day.
Now I use the app in every city I go to, even in my hometown, where I'm discovering new things I didn't know about.
Feel free to ask any question, I will tell you everything I learned along the way.