Transcription
Google AI Studio has just introduced VIP coding, and it's absolutely incredible. You can now build custom AI features right into your prompts. Things like live API, voice generation, and audio transcription.
It can now also freestyle complete app ideas for you, so you can learn how to prompt it in practice. And it even suggests extra ideas to make your apps even more powerful when coding. And best of all, all of these features are completely free. So, in this video, I'll show you exactly how it works. We'll test it out together, and of course, this channel is called Yori Workflow. So, I'll show you how to build actually useful apps with it, and even add simple back-end automations to your apps. So, without further ado, let's get right into the video.
So, the first thing you're going to do is head over to Google AI Studio, and you should see this VIP coding tool pop up right here, or you can just go straight to the build section. Here, you'll see a list of features you can build into your app. You can click on the ones you want, say live API or video generation, and then just describe your app, and Gemini is going to code an app for you with those features included. Let's go ahead and see how it actually works in practice.
I'm going to head back to the prompt section and click "I'm feeling lucky" here. And that's going to generate a random app idea and select the best features by itself. Let's go with this idea. As you can see, it already generated the prompt for me and added "nano banana powered app feature." And the prompt says, "Create an AI headshot photographer app. Users upload a casual selfie. They then select desired style and the app generates a generic professional headshot based on that style." Send the prompt in, and it's going to go ahead and start coding. And on the right, you'll see here that it's now suggesting ideas to improve the app further. You can click through them and select the one you want. I'm going to click to add this one. And as you can see here, it automatically generates the prompt on the left. I'm going to wait for the app to finish development. Send this prompt in. And once the changes are done, here is our app. Let's go ahead and add any image. As you can see here, I can choose the background. Let's say natural setting. Choose soft lighting, approachable, whatever that means. And in a couple of minutes, here is my headshot. I mean, it's not perfect, but the app definitely works. Just like that, you can create a bot that generates custom posts, banners, or edits images for you completely for free.
Now that you understand how it works, let's go ahead and create our own custom app. Now, I'd love to translate my videos into multiple languages, but dubbing tools are usually quite expensive and clunky. Well, let's create an app that fixes that. I'm first going to go ahead and choose a "generate speech" and "transcribe audio" features. And in my prompt, I'm going to say, "User first uploads an audio recording. You will transcribe it in English." Once transcribed, display a drop-down menu with an option to choose from 50 languages and a translate button. Once languages are selected and the button is pressed, translate the transcript into the languages chosen with the most similar amount of characters, tone, and style. Create a tab for each language with the corresponding transcript inside it. Next to each transcript, display "generate voice over" button. Once pressed, it generates the audio in a female voice in that language and make that audio downloadable. Going to go ahead and send that in. And literally in one prompt, my app is fully functional.
I'm going to first upload the original audio for this video. It's about 8 minutes so far. I'm going to go ahead and press transcribe. It's going to take a few minutes, but the quality is really good. So once the audio is transcribed, I can now translate it into any language. Let's go ahead and choose Chinese. Hit translate. Again, this is going to take a minute. My transcript is now translated to Chinese. And at the bottom, you'll see that there's the "generate voice over" button. Let's go ahead and hit that. Now, this took about 5 minutes, but apparently here's how I sound like in Chinese. "Google AI Studio." I can now download that and add it to my video. And you can see that it's about the same length. Now, let's go ahead and do the same with German and Spanish at the same time. And after a while, here's my audio dubbed into German. "Google AI Studios." And here I am in Spanish. "Google AI Studio." Now, if you speak any of those languages, let us know in the comments if it's actually any good. I was surprised by how well this prompt worked and how high quality the audio is.
It is the Halloween season, and you might not be scared of ghosts, but I'm sure you're scared of AI replacing your job. Well, it doesn't have to be that way because my partners at Alskill, the world's first AI-focused education platform, are hosting a two-day AI Mastermind workshop next weekend. This training usually costs about $400, but you can attend it completely for free because I've secured 1,000 free seats for you guys in the next 48 hours as part of the Halloween sale. In just two days, you'll learn to master 10+ AI-powered tools, become a pro at using AI in Excel, Sheets, and presentations, and build AI agents, automating your workflow, and so much more. So, if you're looking for hands-on AI training from practitioners at Microsoft and Nvidia, people who have actually built and shaped the industry, this training will be happening next Saturday and Sunday from 10:00 a.m. to 7:00 p.m. Eastern time on both days. So, go ahead and secure your limited seat in the description or scan this QR code and let's get back to the video.
I create loads of step-by-step instructions for my clients, and it would be amazing to just have an app where I can upload my screen recording and it will automatically turn it into an easy-to-follow guide. So, for my next app, I'm going to say, "User uploads a video, generate a detailed step-by-step documentation from it." And I'm going to add a "video understanding" feature. Send that in. And here's the first app it generated. I'm going to go ahead and upload this video about creating a Firestore service. This video is just a step-by-step tutorial without a voiceover. So, it's just going to go ahead and start processing it. And in a couple of minutes, here's my documentation. Now, it's absolutely fantastic and very detailed. We just need to fix the formatting by removing the asterisks and add text editing for me to polish it. So, I'm just going to say, "Do not show asterisks in the UI. Format the text without showing them. Make the final documentation text editable." Now that it's done, I'm going to upload another video here about getting the API key. And now the documentation looks much better without the asterisks. And here I can click "edit" and remove this intro, for example. Save it. And now it looks perfect.
You can also add any backend to your apps by just creating simple Make.com scenarios. It's completely free and only takes a couple of minutes. For example, let's create an automation that will save the generated guides as Google Docs in my Drive. You're going to first head over to Make.com with my link in the description. You'll get 1,000 free credits a month. This app only needs about three to run it. Then you're just going to head over to scenarios and click to create a new one. You're going to create a webhook as your first step. Custom webhook. And in advanced settings, you're going to add a data structure. You're going to create two fields here: "documentation" for the article we generate, and "title" for the guide file name. You're going to save it. And for the next step, you're going to add Google Docs, "create a document" step. In the file name, you're going to choose the title we got from the webhook. And in the content, choose the document and save it. And for the final step, you're just going to add a webhook response with the default settings and save.
Now, you're just going to go ahead and copy the webhook URL we got from the first step, paste it into Google AI Studio, and say, "Once documentation is created, add 'Save to Google Docs' button. When it is clicked, activate this webhook. Set title, which is the documentation title, e.g., 'How to create a Google Cloud service,' and documentation text, e.g., 'This guide will walk you through blah blah.'" Now, once you receive a 200 response, display "Saved" in the button. Once the changes are made, let's generate a documentation. And you'll now see this button at the bottom to save it to Google Docs. You're going to click "Run once" in Make. And just like magic, your Google Doc is created in your Drive automatically. Once everything is working correctly, you're going to press to activate the scenario. And now this is going to work for literally every guide.
Now, I don't know about you, but I absolutely hate listening to long voice messages, especially when they're very detailed. You have to note everything down. So for my first step, I'm going to select "audio transcription." Think more for complex questions. In an AI chatbot feature. And for the prompt, I'll say, "User uploads a voice memo recording. It gets transcribed. And in a separate tab, the user can ask questions about it. And use the think mode for more advanced questions." I'm going to send that in. I quite like this idea Gemini suggested, so I'm going to go ahead and add it to my prompt. It says, "Generate a basic audio player for the transcribed voice memo. The player should allow the user to play, pause, and seek through the audio. Synchronize the playback with the text transcript so that the current word or sentence being spoken is highlighted." I'll send this prompt in. And once the app is finished edits, here is the first version. I mean, it looks good, but I don't want to record audio. I want to upload it. So, I'm going to say, "Let the user upload the audio instead of recording it." And after the changes, here's what the app looks like.
I'll go ahead and upload this super detailed voice memo detailing property sale conditions. And as you can see, it's now transcribing this audio. And in a few moments, we get the audio transcribed. And if I play it, you'll see that it now syncs the voice with the transcript, highlighting what my client is saying in blue, which is pretty cool. Now, let's turn on the think mode. And I'm just going to ask the AI chatbot to give me a detailed list of conditions the clients listed. And it's going to think longer to provide me with a detailed answer. And here it is. It looks pretty cool. It laid out everything the customer mentioned: the pricing, the conditions, and included items. Now, this is amazing. You can create a bot that analyzes any other audio like lectures, meeting recordings, or even just record the audio yourself and then add any other AI features like an AI chatbot.
So, that's it for today's video. I hope you learned something new and can now build custom apps to streamline your workflow. I would absolutely love to make another video showing you how to build apps with more advanced backend. So, if you'd like to see that, make sure to comment it down below. Thank you so much for watching, and I'll see you in the next one.