Learning from my failures: "I built something fast, but I didn't become the person who could stand behind it."
It all started 4 months ago, when I went out of control and chose to become unemployed. I gave myself 3 months to find out what I wanted to do going forward (I knew I needed longer but I promised the period because my mom was watching). The change of routine from the regular 9-to-5 to having full control of my time back excited me with a fresh shot of adrenaline; I was open to all the possibilities in the world, and I believed I could be successful.
Everywhere I looked, people were telling me how easy it is to build a SaaS (Software as a Service) with AI now.
Believing in its power and my dedication, I started the journey of software development — after a brief learning session with two videos by SuperSimpleDev on YouTube:
With — You, On Earth.
Shortly after, I began my first PWA (Progressive Web App) project 'With' —a poetic set of 366 self-reflection prompts, structured as introspective 'Truth' or 'Dare' cue cards, guiding users through a year-long ritual of personal clarity and renewal.
I attempted vanilla coding with my brand new knowledge but quickly turned to AI to speed the process up because what I really cared about was getting something out there to sell.
It didn't sell.
The lack of public interest in the first product had led me to dive into the next project instantly: my first and last SaaS product which I named 'This, Is' — a Notion-powered site builder that turns your Notion content into a customizable, personal web presence. The idea started merely from the desire to have freely moving 'stickers' on my Notion, and to have a platform where people share their lives with data.
This, Is — A Notion-powered link page that acts like a home for visitors.
I wanted to make something great, fast — to outrun all reasonable doubts.
By the time I felt comfortable sharing the idea, I had already rushed too far to step back.
The idea didn't hold up when I tried to explain it out loud to my sister. It came out like a bad pitch with 'powered by Notion', 'customizable site', 'sharing lives', 'journaled data'… I couldn't introduce my product in just one sentence. It was like I was trying to convince myself along as I watched her expression change.
By the end of my elaborations, she asked questions that hit exactly where I was uncertain:
'If it's for sites, why not Wix?'
'If it's for sharing, why not Instagram, Threads, or X?'
'If it's by Notion, why not stay Notion?'
A slight dejection felt like a K.O. punch.
I thought I had found a market gap, but I had neither the clarity nor the foundation to stand on. The progress after that felt like walking along a crumbling edge. The realization came too late.
I must admit that I had the urge to quit, but I had also set a stubborn goal to finish it by the end of March (later April).
Uneasily, I continued the project, but reshaping it into a more flexible alternative to Linktree — where, instead of a stack of links, your link page could be presented with more freedom and creativity that could feel more like you.
I updated the copy, kept jumping back and forth between Auth, APIs, Routes, Databases, UI/UX… All through AI. It was all fun and games until the structures grew complex. I spent hours debugging under the instruction of Gemini and ChatGPT like a marionette, too exhausted to ask 'why'.
The more I relied on AI, the less I tried to think for myself.
By the end of March, the product was almost ready so I had started to ask AI about soft-launching it, calculating the suitable price range, things I should've asked before starting the project. As someone who worked in management, I find it hilarious how I disregard the planning stage.
Anyway, this was when I learned about the legal requirements of launching a SaaS product — the need for a business entity, privacy notice and terms, then the business part of it — the cost, revenue, the consideration of churn rate, yada-yada-yada.
These procedures were like the bumps on the road. I sped all the way and then I was suddenly forced to slow down. I was glad that my hesitation, which was growing subconsciously, had driven me to take the pause, it gave me time to reconsider. It went from 'should I do it with a sole proprietorship or a Sendirian Berhad' to 'do I really want this?' — I didn't want to quit midway but I was considering quitting when the prototype was ready?
Notice how it went from a 'product' to a 'prototype'? That was an indicator that I humbled myself when 'This, Is' stopped being a project and started becoming a responsibility. I was so ready to sell but I changed it to the 'waiting list' because it became something I wasn't ready to take responsibility for.
Eventually, here came the moment when I finally admitted that I could build something with AI, quickly, but I didn't become the person who could stand behind it.
AI gave me the speed; it fed my impatience as well as my greed and delusions. The more I relied on AI, the less grounded I felt. Working with AI didn't feel like collaborating with people. It felt like commissioning a custom-made Pandora's box.
People say you should sell what you build.
I think you should only sell what you believe in, and I didn't believe in this enough to launch it.
So, yes, I was suddenly ready to quit. I felt relieved to walk away — not just because it was easy (frankly), but because I knew this wasn't something I was willing to carry forward.
I don't just want to build fast — I want to understand what I'm building.
And that means learning to slow down.
I am sharing my prototypes here, feel free to reach out if you're interested:
https://this-comma-is.vercel.app/
https://with-you-on-earth.vercel.app/
