Progress:
My Mvp consists of two main parts. The figma drawing as well as a prompted AI model. I inititally intended to complete the entire app but later found that to be impossible with my skill and time alloted to devlop the entire app.
Here’s My Figma model:
The basic idea of the app is to give investors and easy way to organize their real estate assets. Much like how you can view your stocks on your phone, investors should be able to look at their properties on their phone.
The AI is a prompted version of ChatGPT 4o.
When presenting the AI, I would gather and organize my observations of how AI responds to different prompts. Then pulling out the best prompts to ask to make sure there are no risks involved when it comes to following advice.
Challenges:
I pretty much entered Fusion with zero skill in any stem field. This left me with limited options when it comes to my passion project. The immediate challenge when I started this project was I didn’t know how to build an app. There were websites such as bubble that could help with building a mvp but frankly with zero experience it’s quite hard to figure it out. The next option was to map it out which is what I did. Again, I’ve never used figma before but I assumed it was the easier then building an app on bubble. So that’s what I did. I took a lot of time to learn everything and was extremely frustrated at times because I was not a designer too. By now the obstacles are pretty much the same, it’s going to take a while for me to learn because I’ve never touched a lot of these tools before.
After wrapping up the figma model I had no idea how to approach the AI chatbot. Firstly, the AI can’t be an actual financial advisor because of liability issues and since I wasn’t building the app, I needed to experiment with the AI in another software. Aaron, who was building an AI project of his own, showed me how to model and prompt ChatGPT. Since he spent the time watching youtube and learning the code I luckily didn’t need to do the same. The last challenge I’ve encountered is, how am I going to present this chatbot? Firstly, it’s literally ChatGPT and there wasn’t really a visible difference. I figured the way to present my learnings from prompting the AI would be to run scenarios and common prompts that investors using my software might ask. I would then gather the data and filter out the rubbish or risky answers.
Next Steps:
Moving forward, I’ve asked a couple of my connections to see if anyone would be willing to either help me build the app, lend me the funding the build the app, or to just sell off the project. The figma deisgn would obviously be integrated into the app and the chatbot would likely require it’s own data since ChatGPT can’t be commericially used. The AI is the main problem as that would be the most resource and time intensive part of the project should I continue working on it. I happen to know the founder of an AI real estate start up so perhaps there can be some collaborations happening there.
Leave a Reply