Self Affirm Online

Landing Page
Landing Page
Landing Page
Landing Page
Landing Page
Landing Page
Search page
Just for you page
Plans page
Setting Page
Create page
Affirmation share page

Inspiration

I'm a full-stack developer with three years of informal experience, driven by a deep passion for using technology to make a real difference. This project was born from a desire to help people reconnect with themselves, build confidence, and take meaningful steps toward becoming the best version of who they are - mentally, emotionally, and personally.

What it does

This website provides positive affirmations across a range of subjects. Users can listen to specific affirmations based on what they search through our extensive database. Users can even share affirmations via links with friends and family or create their own custom affirmations.

How we built it

This project was entirely built with Bolt. Normally, when programming, I use Visual Studio Code, but it was not needed for this project. Everything was achieved within Bolt's interface.

The project uses React, Tailwind CSS, and TypeScript for the front end, and Supabase for the back end. It integrates several APIs, such as Groq Cloud for our LLMs, a public inspirational quotes API, and Eleven Labs for the audio affirmation voice-over.

The Images on the home page were generated using Sora AI.

I also created a program in Go to interface with a Supabase Edge function, allowing me to add affirmations through a Discord bot.

Challenges we ran into

Many of the technologies in this project were new to me. While I have experience with NoSQL databases like MongoDB, this was my first proper exposure to SQL. It was also my first experience handling user authentication via Supabase.

Relying on AI to write code can be effective; however, I needed to deeply understand the project’s technology to debug problems that AI could not solve. One recurring issue was that AI is very confident when writing code and often rewrites perfectly optimized code to fix unrelated problems. This caused setbacks because I had to debug issues myself using Chrome DevTools and breakpoints.

At some points some of the components generated with bolt weren't functioning as expected, hence I utilized external AI tools such as chat GPT to generate code snippets to fix those components.

Accomplishments we're proud of

Overall, I'm proud to have created a fully functioning service with a working Stripe payment system. I'm also proud of the project’s pace. What normally would take months was completed in a fraction of the time, which is very empowering.

What we learned

This project exposed me to several new tech stacks and integrations, including:

Stripe
Supabase
Google Authentication
SQL
Tailwind CSS

It also taught me the importance of not relying solely on AI. Having prior knowledge significantly speeds up development and debugging. It’s essential to keep improving our skills so AI can serve as a development enhancement, not a replacement.