Tech stack for AnimeAI (acquired) - How/Why and What's next!
Howdy, fellow Indie hackers! I'm Stacy. And since 2022, I've been an indie hacker and a digital nomad! My first-born AI project, Animeai.lol, found a new home this year. (It got successfully acquired, yep!)
Now, I've embarked on a fresh journey with a new venture, customqrai.com.
I receive numerous questions through Twitter DMs about my project - what tools I used, how I trained the AI model, and the costs involved. Today, I want to answer those questions and share my approach.
Despite a common belief that launching an AI project requires top-tier AI expertise, my experience shows otherwise. Let me debunk this myth - I'm a living, breathing counterexample. The real deal is not the fancy AI wizardry, but finding an idea that fits the market like a glove and executing it seamlessly.
Tech stack:
My background as a backend software engineer greatly influenced the tech stack I used.
So, when I decided to embark on my indie journey, I chose to stick with a tech stack that was familiar to me
Backend: Java + Spring Boot
I know, I know - it might sound as wild as pineapple on pizza to some, but here's why it worked for me:
- Familiarity: This tech stack is something I've frequently used, which allowed quick development and efficient handling of changes.
- Versatility: The Spring ecosystem, particularly Spring Data, provided easy data extraction from the database.
- Code Recycling: I utilized code snippets from past projects, which helped expedite the development process.
DB: PostgreSQL
For my database, I used PostgreSQL for three simple reasons: I knew it, it played nicely with Spring, and there's a sea of resources about it online.
Also the project is not that big to spend too much time thinking about db optimization, I just needed a simple sql db, that’s it :)
Frontend: REACT?!
Buckle up, because here's where I messed up a bit. I went with pure React - a fantastic framework, but not the best fit for a landing site.
Struggling with SEO tags and missing out on the ease and optimizations that come with Next.js made me realize my mistake. So, a piece of friendly advice - give Next.js a shot!
AI: 3rd party API
When it came to AI, I opted for a third-party API. Training AI models can be resource-intensive, and given my limited expertise in this area, using an API allowed me to quickly validate my idea.
I found a popular 3rd part stable duffusion provider and then dedicated time to adjusting the AI and developing the backend communication channels.
There were even times when I reached out to the API developers!
That's a wrap
And there you have it - my tech stack in a nutshell! I picked tools that were familiar to me and let me bring my idea to the market in no time.
Now, I'm weaving a new AI project, CustomQR AI, and spilling the beans on every step of the journey on my Twitter - check out how I do marketing, optimizing my new app and many more!
Feel free to tag along, and remember, keep hacking and stay indie!
40 open-source gems to replace your SaaS subscriptions 🔥 🚀
After 19,314 lines of code, i'm shutting down my project
Need feedback for my product.
We are live on Product Hunt
Don't be a Jerk. Use this Tip Calculator.
Great work! If you don't mind me asking, what stable diffusion provider did you use?
sure, just dm me on twitter
Great, Congratulations 👏 & Good Luck with your next project 🌸 Would you mind sharing your process of selling it ? Like how did you find a buyer and if your site was registered as a company then how?
Thank you!
I published a long tweet with selling process on Twitter, you can find it here
https://twitter.com/stacy_siz
Found buyer on acquire dot com
I'm registered as individual entrepreneur
Hope that helps!
Thanks a lot for sharing it , i would check it out :)
For those wondering what the acquisition price was:
Wow, I didn't expect to see this number. I thought that promising projects cost much more. Either way, it's still a great achievement. Good luck in your future endeavours :)
Was going to suggest using nextjs but then I realized you wrote about it. Try to take advantage of server side rendering and static side generation. That what boost your website to perform better on SEO.
For reference check nextjs documentation: Rendering.
yeah, nextjs is simple and fast :)
Role model for learning, keep going
Thanks, man! Means a lot to me 🫡
that's so cool! I also have a project on React and struggling with SEO. Is there any other way or NEXT.js is the only way?
There is certainly a way!
For anime ai it magically hit the good traffic, so there is a way :D
Btw I posted traffic updates on my Twitter if you want to check out https://twitter.com/stacy_siz
I tried to use react-helmet lib to make everything right with seo headers
But this lib just didn't work on my project somehow
Maybe you would have a better luck :D
I must say that anime ai website had 3 pages, so there wasn't much content for seo
What helped me personally was domain name + Product Hunt launch
Because of PH it boosted my website on 1st page of Google :)
Also time, seo is a bit slow
tried react-helmet as well and didn't work for me either, maybe its time for NEXT.js already:D
how did you get your first website visitors/clients? we just launched our SaaS product and struggling to get any visits. Have you used only Product Hunt or other things as well?
Maybe nextjs is the way then :D
All my first visitors were folks from Twitter as I run build in public blog, people were interested in checking my tool out :)
Congratulations! Your technology stack choice is almost the same as what I have been considering recently. For the front-end, I chose Next.js. However, I don't have any ideas yet for what product to build.
Thank you, Lyon!
Yeah, java + spring are heavy army stack and at the same time reliable and flexible :)
As for product idea I can recommend Reddit + Twitter, check what people are talking about and building
Good luck!
Thanks for your honesty about struggling with a pure React landing page. I feel this on such a personal level. I'm from the VueJS world, and my 1st and oldest micro SaaS, Displagent, is still a pure VueJS website. After a year, I hardly have any SEO gains to show from it.
I decided to use Nuxt for my 2nd product, a C# job orchestration platform called Didact, where I use state site rendering for hyper-optimized SEO. It's already looking 10x better on Google than Displagent does.
I'll be changing my Displagent site to Nuxt very soon.
Pure React, Vue, and Angular are great tools for traditional SPAs, but for landing pages, like you said: you need something more like NextJS or Nuxt.
Congratulations on the acquisition! I like that you used a framework you’re familiar with (Spring)! I see people often get distracted by the latest shiny new framework & this can dramatically reduce speed/agility.
I’m curious regarding your mention of using Next.js, do you feel as though the benefits are only for the landing page, or for your whole site?
Thanks! I'm not a big expert in frontend tbh
Also depends on what do you mean by the "whole site"
If this is a website with simple pages like: about, contact us, gallery, features etc
I still think nextjs is a great and better choice :)
I'm not either! :) I have used Next.js before and I like it, I feel like it's good for static sites but I'm a bit apprehensive about using it for a dynamic web app.
For example, I wrote a simple app that integrates with ChatGPT and Next was perfect for that, but maybe if I were doing something more sophisticated I might just use pure React.
I agree with you though, I would probably also use Next.js for what you described there!
Congrats
Thanks :)
Hi Stacy, first of all congrats on your achievement! I'm pretty new to working on side project like you guys working on, there are several dummy questions on my head:
In a nutshell, I just want to cut the cost as low as possible to prototype first. Would be very grateful to get feedback from you!
Glad to see you share.
Thank you! Let me know what are other topics you're interested in for the article :)
Hello friend I seriously read your article, I saw a while ago and you similar to the QR code about the artificial intelligence of the creative products, although I have not experienced the two products, but the current market environment every day there will be more creative generation, you can more face the market, listen to the user's feedback, I believe that for your product will be of great help!
I'm not sure, but I think it's animeai. app now.
Nope, that's the other app
Mine was animeai dot lol
Love to see this. Thank you for the post, was very helpful. And provides the validation I needed to look into next.js.
Thank you and good luck with your app and nextjs experience :)
soo cool, love the idea!
Thank you -
Congratulations,Stacy more wins and thanks for sharing your process too!
Thank you! Let me know if you have any other ideas on what would be valuable for the next article :)
Thank you for sharing! !
Always welcome 😎
Hi Giga Stacy, I tried visiting animeai.lol but it's down! Did the buyer close this one?
Hi! I hope website will be live soon, have not so much info about it myself!
$300 ARR?
So you skipped the marketing step, it would seem.
Building the product is the easy part. Selling it ... the harder part.
Yeah, marketing is tough ...
Congrats for selling it, but building your marketing skills is just as important - if not more - as training your bot.
Here's a Challenge for your next startup: try to push your sales to $100k ARR ... so that you can sell the whole startup/venture for $1M !!
(my formula to get easy valuation idea: 10x ARR)
It might well be a strategy to build a personal brand: quickly slap together some crap, earn $300 with it, sell, post about the sale on all social media, repeat.
If you check her replies (including tweets), it's all "dm me on twitter", "dm me", "dm me". Just marketing and pumping up the twitter account, not a real sharing of business insights.
wow, didn't expect these kind of replies
"slap together some crap"
First of all, I had people who used service both for free and paid and they enjoyed it. They used anime pics as their profiles etc
So, if this seems like "crap" to you, for me and other people it's not
I think if you judge people and their projects like this, you won't achieve great results not in business not in relationships
Think about changing your mindset
"dm me"
You do not need to follow me to be able to dm me, my dms are open
And I'm happy to talk to people and share more details in personal conversation
"not a real sharing of business insights"
In my Twitter I share what tools I use to build product, how do I market it, what challenges I encounter, I also spark conversations for example when I need to do something I haven't tried yet eg: I was choosing tool to email people on whaitlist and founders in comments shared their opinions, this was helpful
Again lots of people send me messages that in my blog I genuinely share value and it helped them personally
And finally my target audience is not even on Twitter
I communicate with founders there
But you Jeff Hoffman just created your account yesterday to post ugly comments under my post
Okay 👌
Well, that's a lot of effort deployed for not much to show for it, if that's the case.
(and I'm never on Twitter, sorry Elon)
The point of having a brand, ultimately, is to sell stuff, as the brand acts as a shortcut for whatever you position it to be.
But in order to sell anything, you must provide "real" value ... no brand can save you if it does not provide serious value first and foremost.
Value is the key !
If no value under the brand, what's the point ...
Nice work @StacySiz 👏 I'm building mine with a more "JS-oriented" stack, I will share the result later!
Thank you! What's your project about?
Simple AI assistant, but with a prompt writing helper
Congratulations! So inspiring to see awesome AI projects like this being released into the world. Can't wait to see more from you ;)
Thanks a lot for the warm words 🥹🤝
Nice work
Thanks mate 😎
Nice
Thanks man 😏🤝
Congrats with the acquire. I really curious with your marketing on twitter. Could you share your twitter account. I really want to follow and learn from you ;)
Hi and thank you! Sure, here is my Twitter
https://twitter.com/stacy_siz
Let's connect :)
Since the site is no longer up and live, could you tell me what the site's purpose was?
Also, I'm curious about how you found a buyer and how you came up with a price.
And congratulations on selling your app!
Here is a Product Hunt page to help with visuals
https://www.producthunt.com/products/animeai-become-anime-heroes
Basically:
You can see the anime results on PH :)
As for buyer - I used https://acquire.com/
That's so rad! Thanks for sharing that with me. I love those images it created.
At least the last time I checked you couldn't upload sample images into DALL-E for it to build off of. I'm sure there are a bunch of great AI image tools out there though.
Do you mind sharing what AI API tool you used for the image creation?
Hey Stacy, how long did it take for you to build animeai.lol and how many figures was it acquired for (5,6,7, etc)?
Hi, Justin!
It took a few weeks but I was adding some things over time
Also the longest part was to get approve from payment system, verification process took a month
As for figures I posted tweet about the whole story if you're interested in reading :)
I would love to read it! Unfortunately, I don't have a Twitter account and I can't remember if you need an account to read tweets now😅
Congrats on selling your business! Curious... Did you build it with the intention of selling? Did any part of you want to hang on to it?
Thank you!
To be honest I didn't have any selling intentions
It was my first solo shipped project and I also wanted to experiment with AI :)
Then later growth was a bit slow and I needed to focus on my other SaaS, so I selled it!
That's awesome! Congrats!
I thought about how something like this could be offerred as a saas from the first time I saw those QR codes videos on IG. I'm not a coder, so I wouldn't know how to build it. But I think it definitely has great potential! All the best 🙌
Thank you!
Yes, I also feel like CustomQR AI has a big potential, right now cool marketing is needed, so I'm trying to figure out this distribution part
Any ideas? :D
Not completely sure. I'd try to think in the value proposition, since a QR is not something that makes money by itself, then you have to find a way to sell it basically. Like, maybe for people wanting to display ads on street screens and such. Because for social ads it wouldn't make sense because they wouldn't re-scan the screen of their phones.
Love the simplicity! And the Custom QR Codes look sick.
Thank you!! 🥹
Yesss, I'm setting high quality standard for CustomQR AI :)
Now the key is distribution heh
Java and Spring are very solid even though they get a lot of hate for being old. Most of the Fortune 100 companies all have important Java infrastructure of some sort; so kudos for making it work for your use case.
Thank you!
Yeah, I think they are getting hate because they are heavy and has looots of stuff that easy project do not need, also java is long and strict (unlike python)
That's why serious projects from Fortune 100 use serious java stack :D
For me dev speed was important, so I took stack that I know the best :)
Where is the site live? Animeai.lol is dead.
I'm guessing maybe it was taken down after being acquired?
Yes, I hope it soon will be live again
Congrats at getting acquired! Must have been a great journey
Thank you!
Yeah, it was and I learned a lot :)
Glad I documented everything on Twitter heh
Congratulations on getting your AnimeAI acquired! What are your next steps now?
I'm focusing on CustomQR AI now, just launched! + making something else 😏
Woohoo! Rooting for you, StacySiz!
Thank you, man🫡
Congrats.
So it's a simple 1 time purchase website, app sent results on email, so there was no need for auth :)
What helped in this project:
Nice story, thank you for sharing.
Maybe a silly question, I've never sold a product before - in practice when you sell a product how does the 'transfer' happen? i.e. do you just give the buyer access to github repos? What about the hosting? I host most of my stuff on AWS but I'm not entirely sure how I'd go about transferring it to a buyer
I think it depends on a service that you're using, some of them has a feature to transfer. For example it was really easy to transfer domain with Namecheap
As for hosting web app was pretty simple and almost didn't depend on data, so I just wrote an instruction document how to set things app + provided migration scripts
Hope that helps!
This comment was deleted 2 months ago.
Thank you! Glad my learnings are helpful for the community!
I also get tons of inspiration from folks on Twitter :)
This comment was deleted a year ago.