Here’s Charles-app, a native mobile application built using Bubble technology.
Let me know what do you think about and feel free to get in touch if you need help on your projects.
I developed this app in 5 months 
, not including the first version (which was a PWA) or the testing inherent to the product’s design process.
Today, Charles-app 
is available on both the Play Store and the App Store:
Frontend? Bubble Native, officially still in beta, but the app works incredibly well regardless. It would be a lie to say I didn’t use multiple workarounds—innovation often requires pushing boundaries, based on my experience. I didn’t bend the tool; I simply pushed its limits beyond what’s currently possible. I developed around 20 plugins in React Native and JavaScript.
An app like this can’t function without a custom backend. So, I didn’t just build an app—I built an admin-friendly platform with a custom CMS, also created using Bubble.
The microservices ecosystem includes Render, where I host several services and proxies in Python. The goal was to minimize reliance on other no-code solutions like Make or Zapier, to maintain full control over critical tasks.
-
Pinecone for vector embeddings, used for data analysis and storing the knowledge base for the AI agent.
-
Google Gemini and Mistral AI via their APIs: prompt engineering, structured responses, and as a learning tool and super assistant
. -
The AFP API for fact-checked news, plus other confidential APIs for specialized features.
In conclusion:
I learned a lot by setting this ambitious goal.
I learned quickly thanks to AI.
I executed my project and maintained control using no-code, which made my goal achievable.