I validated my product with No-Code, now I'm dropping it for a custom build
I created thisisbuiltwith.ai in November of last year in response to the growing hype around AI.
I wanted a place to discover all the new projects being created with AI, and in the essence of transparency, this would act as the perfect lead gen for when I created an AI product of my own.
The site is an AI product directory with an attached newsletter targeted toward businesses, brands, and builders.
Each week I showcase:
3 new AI products
2 useful tools for businesses
1 idea or API you can use as inspiration to start building with today.
I am a software engineer by trade, and my main skill set is with .NET, but with AI taking off the way it did, I turned to no-code.
Here's my current stack:
Website: Softr.io
Database: Airtable
Newsletter: MailerLite
Softr:
Softr was a fantastic tool to work with, it helped me get my directory out in under one week. I found it intuitive and easy to understand. The issue I have found since, like most no-code tools, there are limitations you have to take into account.
This is the usual trade-off when considering any tech stack. Speed to develop vs Customisation/Cost.
Airtable:
Airtable was a tool I had heard so much about and couldn't wait to get my hands on, and I love it. Quick and easy to use and integrates with Softr perfectly.
Mailerlite:
Mailerlite is a choice I had to make based on what integrated with Softr, and I'm glad I had to. I prefer this newsletter tool to all the others I have used previously. I find they have a very generous free tier and even the paid tiers are Indie hacker friendly. The drag-and-drop interface is very easy to use and adding new sign-ups was ridiculously straightforward because of how easy it was to integrate with Softr.
I've dabbled in no-code before, but not to this extent. It has been a great learning experience, but looking ahead to the future, I've now validated my idea and confirmed the route I want to take the website next... so I've chosen to build it custom.
New stack:
Frontend: React
Backend: .NET Core
Hosting: Heroku
Database: Airtable
Newsletter: Mailerlite
The next phase of my product is being built as we speak. The main reasons are customization as I mentioned earlier, I'm also using this as a learning opportunity for some tech I've wanted to learn for a while, and the product is free, so it's more cost-effective for me to host it myself than I would for $40 a month with Softr.
.NET core was chosen as it's something I have a lot of knowledge of. React because I've wanted to learn it for a while and found this project to be the perfect way to do that, plus, the next product I have in mind is very UX-heavy, so this will fit in perfectly when it comes to building that.
Airtable is coming with me for now, looking forward to seeing how I integrate it with .NET core, and Mailerlite is my choice of newsletter provider from here on out, so that's an obvious choice. Finally, hosting on Heroku because that's where I host all my projects so I don't feel the need to change that yet.
That's the update, stay tuned for This Is Built With AI V2!
Happy to answer any questions about the no-code tools, product, AI, or whatever comes to mind!
Subscriebd. How do you find/search the sites that were created via AI?
A mixture of Indie Hackers, Product Hunt, Reddit, other newsletters and being active in the space on Twitter.
I would love to know the whole process of validating your idea and how you did it. I am wanting to validate an idea with a code script CEX and Alchemy
Also, if you want some free UX / UI design and React help, I would be more than happy to do it. Message me if you are interested!
Hey! It was just a mix of watching subscribers come on board, observing the click-to-open rate on my emails and having conversations with people that reply to the email. Helped me get a feel for what people might need in this space.
Happy to discuss your idea on Twitter - https://twitter.com/JoelPickin
I might have to take you up on that offer sometime!
Hey @JoelPickin,
Really cool project, you have a new subscriber! I especially like the 3-2-1 email format that has worked so well with James Clear ;).
One question I had on the tech stack, it seems like what you're building could be done fairly easily with more 'traditional' website building tools like Wordpress/Shopify/Webflow/etc.
I'm curious if you explored using those or, as you said, you wanted to try out new tools and are using this project to learn?
Ha, you see where I got my inspiration. I usually prefer newsletters that are straight to the point, so I've tried to give it that feel.
It could have yeah, I was thinking of Webflow, but I chose the stack I did to have an easy-ish project to flex my skills before I move onto something that is very UI heavy.
This is amazing! Good luck team!
Thanks Cody!
Great post! We created https://www.codesmashapp.com/ to help make this creative process more continuous and prevent having to start over. Interested in feedback!
Hey @JoelPickin, firstly congrats on such an amazing product, it sounds pretty cool. I'm making my current AI product too, codenull.ai which is a no-code AI model maker. and had the frontend designed in figma but the process to convert figma to react is time consuming and I wanna get the MVP out soon. I have a backend setup in a VPS with an API that does the AI computation.
Do u think there's a free no code tool that can help me connect to custom APIs? Not bubble cuz the learning curve is too high there.
Thanks!
This makes sense to me. No-code tools are amazingly fast at helping you get a quick product out, but when it comes time to up the quality, custom is the way to go. Thanks for sharing it's helpful to see what that process looks like and the reason behind it. Also exciting that you're moving onto V2!!
Appreciate the comments. I totally agree, with a lot of no-code you have your hands tied to their limitations. Maybe that's just my limited knowledge of them though.
super cool and quite the graduation. Any lo-code projects that need more data intensive operations I'd love to recommend patterns.app !!
Haven't heard of that, I'll check it out.
Cool stuff — I actually am in the exact same boat haha (except I’m using a bit of Zapier w/ Airtable and softr).
Coming from a self-taught dev background some stuff was definitely frustrating using no-code, but a lot of things were definitely easier.
Out of curiosity, how did you manage each “article” / post? It’s hard to host any sort of blog using softr and Airtable (especially with rich text), so I’m using a convoluted method of Wordpress -> RSS -> Airtable, and reading each post as code, if that makes sense.
Nice one! I found my previous dev skills came in handy as you have a high-level view of how a lot of the integrations worked.
To answer your question, I didn't. One of the reasons I'm rebuilding is so I can create a blog using GraphQL and get some SEO benefits from it. I do know there is this article which covers how to do it until Softr release their blog functionality.
https://docs.softr.io/blogs/cQCt7G8nr4vgzZuCsrHL6X/how-to-create-a-blog/5Vhtsqxp1VFpURgjGbdVVc
very interesting -- good to hear that I'm not alone in quickly running into some limitations with no-code :)
Hey @joel Can I ask how you come to conclusion about idea validation?
I got my first 100 subscribers, then sent an email to some asking questions about them, their industry, and what they are looking for in the AI space.
This has influenced the new direction with the site.
hey @joel pickin great post - I was curious whether you considered a "higher end" low-code tool like say Mendix (https://www.mendix.com/) instead of full blow React/.NET etc? I'm not affiliated or anything like that with them I'm just curious about whether something like Mendix is a good enough improvement over Softr? Or could you say more about why you don't like Softr?
I hadn't actually, I've used OutSystems which is a similar tool in my day job previously, but one of the main reasons for the rebuild was to get to grips with a frontend framework for the next project I build, it's UX heavy and I'm really excited about it!
I like Softr, I just have the ability to do something better custom, I'm sure an experienced no-code dev could adapt Softr and Zapier etc to do the same I will be doing custom. Plus, I wanted to save $40 per month of hosting the site by doing it myself.
Any difference from futurepedia.io?
Currently no, the new site will be. Futurepedia is focused more on everything, I want to dig deeper into only the best for each business use case.
My newsletter is different. I showcase 3 new products, 2 useful business tools, and 1 idea or API for inspiration for builders who want to get into the space.
I dig into the products and write the newsletter myself.
I have experienced a similar timeline as well. No code is great for basic CRUD/display information-type apps, but falls short on customization of flows, etc. I even find that no code challenges me as a developer to focus on the fundamentals again (aka spend time on the backend, API, etc vs the UI side of the house).
You've nailed it, that's exactly the reason for the switch. Plus I needed to use it as a learning opportunity for my next project which is UX heavy, and it will be nice to save $40 a month on hosting costs.