Created with GPT 4o
I’ve always been fascinated by how AI can understand the world, especially nature.
So, I asked myself: Can I build an app that identifies any plant, without writing code?
Turns out… yes. All I needed was Gemini and Firebase Studio.
The Portal: Firebase Studio
I opened Firebase Studio. You can code here — but I didn’t.
Instead, I used the magic door: Prompting.
You can start from a template, a repo, or… just describe what you want.
SS of the output
But instead of using codes, let’s do it with the prompt. I always wonder how apps can calculate calories from a given photo, so let’s develop one, by just prompting.
Here is my prompt:
An app that tracks calorie of a given photo by first creating ingredients
Here is the first draft.
Good, let’s click on Prototype this App.
Good, it started creating this app. Amazing.
So fast and No-code!
And now the app is ready in 5 minutes.
CalorieSnap
It asks you one thing.
Gemini API.
Collect your Gemini API from here, because you had to put here — Gemini key update.
Testing Time
LeafyAI
I picked a plant image.
Photo by Yuu on Unsplash — a beautiful hibiscus.
Good, let’s use your plant's image. Here is the one that I’ve found ;
Here is the explanation;
Hibiscus 🌺 is so lovely, no wonder that this flower attracts insects 🐝🐞🐛🦋
Let’s upload this image to our Leafy AI.
Here is the output.
SS of the LeafyAI
In just 5 minutes, without code, we have developed an app.
Ready to Go Live?
You can also publish the agent by just clicking.
You must default the project, link the cloud billing account, and set your environment.
And you need 10 more minutes to deploy your app.
Final Touches
It is good, but let’s make it great.
Click on here and explain which changes you want.
Here is my prompt
Design a sophisticated plant identification interface for LeafyAI with these specifications: Color palette: Use a harmonious gradient of greens (sage, emerald, forest) with subtle complementary earth tones Upload area: Create an elegant drop zone with a thin dashed border and leaf icon, transforming to a solid border on hover Typography: Implement light, modern sans-serif fonts with adequate spacing Button styling: Design a gradient green button with subtle shadow and hover animation Background: Incorporate a very subtle leaf or botanical pattern or texture that doesn't distract Layout: Maintain generous whitespace with perfect alignment and balanced proportions Visual feedback: Add delicate animations for uploading and processing states Micro-interactions: Include subtle hover effects on interactive elements Logo: Refine the LeafyAI logo with a more distinctive leaf element or typography Overall aesthetic: Aim for a premium, clean interface that feels professional yet approachable
Let’s see the final look.
Pretty amazing!
And it is live!
More Edits
I could not stop myself. It was so easy, so I want to add more! Next, I developed a feature where you can ask more questions after LeafyAI outputs.
Here, the app is using the API you gave it first, and here is the result.
After each change, you should click on Publish on the top right.
Publishing
And it will make your changes alive.
LeafyAI
Final Thoughts
In just 15 minutes, your app was made and available on the web. And you did this just by clicking.
One word: Amazing!
Google is coming big! Be prepared for the AI summer, because otherwise, AI can also automate your tasks in the future, and your colleague who adapts to AI can be your boss soon.
Thanks for reading. If you want me to send the AI Builders Playbook, subscribe to me on my Substack here.
“Machine learning is the last invention that humanity will ever need to make.” Nick Bostrom
📧 Stay Updated with AI Insights
Join 10,000+ subscribers getting the latest AI, Data Science, and tech insights delivered to your inbox.
💡 No spam, unsubscribe anytime. We respect your privacy.