Journal

Todo App — My First AI-Built Product

How a small task manager became my first end-to-end lesson in AI-assisted product development.

A clean desktop to-do app interface with tasks, progress and calendar

When I decided to learn AI-assisted development, I knew I needed a project that was small enough to finish but realistic enough to teach me the complete product-building process. A to-do app felt like the perfect starting point.

This was not just about creating another task manager. It was about understanding how AI can become a development partner—from planning and product thinking to writing code, fixing bugs and shipping a working application.

Live demo

Try the app: https://jobishkm.github.io/todo-app/

Why I built this

As a Product Designer, I have spent years designing interfaces and user experiences. Development, however, was always the part that stayed outside my comfort zone.

With tools like ChatGPT and Codex, I wanted to see whether I could bridge that gap and build a real product without having a traditional programming background.

The goal was not perfection—it was learning.

My process

I started by defining the product instead of jumping straight into code.

  • Planned the core user experience
  • Designed the information architecture
  • Created wireframes and interaction ideas
  • Broke the product into small development tasks
  • Used AI to generate, review and improve the code
  • Tested every feature in the browser and refined the experience

Rather than asking AI to build everything in one go, I worked with it step by step, treating it like a collaborative developer.

What I learned

This project completely changed how I think about building products.

I learned how to:

  • Structure a product before writing code
  • Communicate requirements clearly to AI
  • Review and improve AI-generated code
  • Debug issues by working collaboratively with AI
  • Iterate quickly without losing design quality

More importantly, I realised that AI does not replace product thinking—it amplifies it.

Tools used

  • ChatGPT — Product planning, UX guidance, problem solving, debugging and prompt refinement.
  • Codex — Front-end development, implementation and iterative code generation.

Looking ahead

This to-do app is the first milestone in my AI development journey. Every project after this will become more ambitious, but this is the one that proved I could transform an idea into a working product with the help of AI.

Sometimes the smallest projects teach the biggest lessons.

Back to Journal