Here's what you didn't know you needed: Mini-AI Apps from one prompt
Builders, get ready. The way we prototype and launch is changing, and it's happening quietly, with a tool that was right under my noses that I just discovered... a new experimental tool from Google Labs called Opal that let me build, edit and share mini-AI apps using natural language and a visual editor.
I've been talking about the power of using combinations of AI tools to amplify your impact, and this new Google tool is the perfect example of how that's becoming a reality for everyone so decided to create a little tutorial. The idea is to chain together prompts, models, and other tools into a functional app, all with no code required.
This is a game-changer for people who need to test ideas fast or build custom tools to boost their own productivity. It's about moving from a vague conversation with a model to a purposeful, delegated task. Instead of just chatting, you're building a tool that performs a specific function for you over and over again.
Let’s try a use case that will resonate with a lot of you: Automating a "LinkedIn Cover Creator" Mini-App.
Here's a step-by-step tutorial on how you could build this with Opal:
Step 1: Get Started
Navigate to the Opal website, which is https://opal.withgoogle.com/.
You will be prompted to sign in with your Google account. Once you are signed in, you can browse the gallery of pre-made apps or click "Create New" to start your own.
Step 2: Describe Your App
In the "Create New" section, you can build your mini-app in one of two ways:
Conversational Mode: You can simply type in a conversational prompt to describe the app you want to build, and Opal will create it for you. I prefer this one, as you can see I just typed a sentence and it works!
Visual Mode
As you describe the logic, Opal will generate a visual flowchart or "node-based workflow" showing each step. You can also build by manually dragging and dropping different components:
Build the Workflow (The "How-To")
Set the Trigger: The first step is to tell your app when to "wake up." You can instruct the app to start manually or set a scheduled trigger to run at a specific time, like every Wednesday.
Fetch the Data: You can provide your profile URL.
Add AI Analysis: This is the core of the app. Add an AI step to your workflow. You can then write a detailed prompt to instruct the AI on what to do with the data it retrieved. Use a template like this: "Given a set of recent LinkedIn posts, please write a nice report with analysis on what the person is talking about and which of their content is resonating.".
Step 3: Test and Share
Opal lets you preview your app in real-time as you build it.
You can interact with it on the left side of the screen to make sure it's working as intended. Once you're happy with your mini-app, you can instantly share it with others using a link. Here’s the LinkedIn cover generator!
Anyone with a Google account can use the app and their own account for the AI model calls.And there you have it. You've just created a powerful mini-app that automates a task and gives you actionable insights!
No hosting. No deployment, and free! Try it here.
Show me what you’ve built, and I’ll feature your work on my next newsletter!
What’s next?
Join my free NotebookLM masterclass