Question 1

Can Google AI Studio really build a full-stack app with Firebase login and database?

Accepted Answer

Yes—and that’s the whole point of this video. I prompt the front end in AI Studio, then I add Firebase email/password authentication and a Firestore database so users can sign in and get their own saved history. Once you deploy, it behaves like a real app with synced data across devices.

Question 2

How do you add Firebase Authentication to a Google AI Studio app?

Accepted Answer

I create a Firebase project, add a Web app, and copy the Firebase SDK snippet. Then in AI Studio I literally prompt: require login before using the app and implement Firebase email/password auth, and I paste the SDK code. After that, registration and sign-in work (just remember: testing auth doesn’t work properly in preview—deploy it).

Question 3

How do you save YouTube summaries and transcripts to Firestore per user?

Accepted Answer

I create a users collection, then inside each user document I add a videos subcollection. In AI Studio I remove the demo history and prompt it to save each pasted video as a document under users/{uid}/videos, then update that document with insights and transcript when they load. The UI history is just a reflection of those Firestore documents.

Question 4

What Firestore security rules do you use so users only see their own data?

Accepted Answer

I generate rules that allow a signed-in user to read, write, update, and delete only their own user document and their own subcollections. I do it the fast way: I ask AI Studio to produce the rules for my exact collection/subcollection paths, then I paste them into Firestore Rules and publish. You must customize the names (like videos, gmail, youtube, googleCalendar) to match your database.

Question 5

How do you generate the YouTube summary and transcript with Make.com and Gemini?

Accepted Answer

I build two Make scenarios: each starts with a custom webhook that receives the YouTube URL. Then I use the Gemini “Generate response” module (with an API key from AI Studio) and pass the URL as a file URL, with simple system instructions for either insights or transcript. The last step is a webhook response that sends the text back to AI Studio to display in the UI.

Question 6

Why does the transcript sometimes fail with JSON parsing, and how do you fix it?

Accepted Answer

Gemini usually returns a clean JSON array, but sometimes it wraps it in a ```json code block even when you tell it not to. I handle it by telling AI Studio to ignore the code block and start parsing from the first square bracket. And if parsing still fails, I add a “failed to load / try again” flow that re-triggers the same webhook.

Question 7

Why can’t I watch embedded YouTube videos inside Google AI Studio preview?

Accepted Answer

That’s normal—you’ll see an embed error in preview mode. I show this in the video: you need to deploy the app (rocket icon) to actually watch YouTube embeds inside the app. After deployment, the embeds work as expected.

Question 8

How do you build a notification hub that pulls Gmail, Calendar, and YouTube into one app?

Accepted Answer

I prompt the dashboard UI in AI Studio first, then I set up Firestore with three subcollections under each user: gmail, youtube, and googleCalendar. In Make, I create scenarios to watch YouTube search results, Gmail emails by query, and Calendar events, then write each notification into the right Firestore path. Once AI Studio reads those collections, the UI updates automatically—and deleting from the front end can delete in Firestore too.

Google AI Studio Just Unlocked Firebase 🤯 (Login + Database in Seconds)

🛍️ Products Mentioned (3)

Make Product

→ Email: antonina@youraiworkflow.co → WhatsApp:

Let's build your app together

About This Video

Frequently Asked Questions

🎬 More from Your AI Workflow