VirtualSnap

A virtual product photography studio powered by AI

Visit Website
June 8, 2023 I built my MVP during a weekend with the help of AI

Hi everyone!
I'm Fernando, and I've built VirtualSnap (www.virtual-snap.com), an Virtual Product Photography Studio which will soon be launching on Product Hunt.
I've been working on this product for almost 3 months now (on and off, since I have a 9-5 and freelance writing work already) but I built the entire MVP in a weekend, and I wanted to share the process with you.

What is VirtualSnap though?

Just for context, the product itself is called VirtualSnap (https://www.virtual-snap.com) and it's meant to be an AI-powered Virtual Product Photography Studio.
In other words, you can take pictures of your products without having to hire a professional photographer.

Cool right? At least I think so!

How I built the MVP

To built VirtualSnap's MVP, what I did was make extensive use of ChatGPT. Yes, shocking, I know.
I'm building it using NextJS, so it's all JavaScript (React), and ChatGPT is pretty good at generating JS code, so it was relatively easy.

I started asking for the front-end's layout code, styled with Bootstrap (because that's all I know), and once I had the main UI, I started working on the back-end.

For the back-end I was using an external API to run the Stable Difussion models, so I tried to be as specific as possible. I would ask for things like:

Give me the code for a NextJS API endpoint that receives a text prompt, a user ID and uses them to send a request to an external API. The response from the API save it to a mongo database, and grab the URL parameter and upload it to an S3 bucket called "photos"

That's the key with chatGPT, being specific. So with this, you get code that's almost ready to work. I only had to make a few tweaks, add the proper secret keys and API keys, and the whole thing would be working in less than 10 minutes.

What about the graphical aspects of the app?

Well, I left that to chatGPT as well. Clearly that's not a great idea, but at least it gave me a non-ugly color palette that I could use across the app.
I've been tweaking the styles manually since then, but the main look & feel has been kept.
And honestly? That helped me save hours (probably days) of thinking how to make it look half as good.

And on top of that, I also used AI to generate the app's logo. For that I went with LeonardoAI. It's a great tool that can generate high-quality images (much like Midjourney) and for the time being, their FREE version is actually useful.

In the end, I ended up releasing my MVP by Sunday afternoon (having started working on it early Saturday). Since then, the project has evolved and new features have been added, so now it's time for the big launch: Product Hunt.

If you'd like to show your support (I would greatly appreciate it), you can check out the ProductHunt page here (the product goes live on Monday 12th of June): https://www.producthunt.com/products/virtualsnap-virtual-product-photography

Happy making!

Comment

About

Generative AI can help streamline the process of coming up with new products or taking existing ones and snapping pictures that never existed (or are too hard to create) for a fraction of the cost.