Souko
From designer to builder: Launching my own time tracking app with AI & no coding experience
From designer to builder: Launching my own time tracking app with AI & no coding experience
(Souko)
(Souko)
Year
Year
Live app
Live app
(Origin)
Existing time trackers felt boring and uninspiring (for me). A disconnect from why I work. I wanted something more than a stopwatch. A tool that made reflection part of the workflow, a space where tracking time could also fuel better habits, decisions, and focus.
(Approach Highlights)
(Opportunity)
Souko started as a personal challenge: what if I could design, build, and ship my own time tracker while learning new AI tools, experimenting with PWAs (a progressive web app), and stress-testing my own design process? Spoiler: I learned more by building than I ever could by planning.
(Approach Highlights)
My process to build Souko
My process to build Souko
The challenge
I set out to build my own time tracker. No coding skills, just curiosity and a personal need. This became a hands-on challenge to figure it out by building, learning, and experimenting in real time.
Rapid Design Sprint
I tested my custom rapid design sprint by creating a 3-screen concept in under 2 hours. I pushed my workflow to turn ideas into prototypes fast, using AI to accelerate design decisions.
Idea Validation
I shared early concepts on Threads and joined the BuildInPublic community. Real user feedback, paired with AI validation, helped me iterate quickly and confirm my idea before writing a single line of code.
Define
I mapped the optimal tech stack for a lean PWA: React, Firebase, Auth0. Then, using Gemini, ChatGPT, and Cursor, I built Souko feature by feature. Debugging, coding, and learning as I went, guided by AI every step.
Design for Performance
Building Souko showed me how design choices impact performance. Every read, write, and feature mattered. I learned to design with data in mind. Optimizing user flows, error handling, and structure to create an app that looked good and ran better.
Alpha & Beta Testing
Friends tested Souko within 7 days, uncovering bugs and data sync issues. After 14 days, 150 beta users joined via Threads. Giving me live insights to improve features, fix problems, and make the app better on the fly.
The challenge
I set out to build my own time tracker. No coding skills, just curiosity and a personal need. This became a hands-on challenge to figure it out by building, learning, and experimenting in real time.
Rapid Design Sprint
I tested my custom rapid design sprint by creating a 3-screen concept in under 2 hours. I pushed my workflow to turn ideas into prototypes fast, using AI to accelerate design decisions.
Idea Validation
I shared early concepts on Threads and joined the BuildInPublic community. Real user feedback, paired with AI validation, helped me iterate quickly and confirm my idea before writing a single line of code.
Define
I mapped the optimal tech stack for a lean PWA: React, Firebase, Auth0. Then, using Gemini, ChatGPT, and Cursor, I built Souko feature by feature. Debugging, coding, and learning as I went, guided by AI every step.
Design for Performance
Building Souko showed me how design choices impact performance. Every read, write, and feature mattered. I learned to design with data in mind. Optimizing user flows, error handling, and structure to create an app that looked good and ran better.
Alpha & Beta Testing
Friends tested Souko within 7 days, uncovering bugs and data sync issues. After 14 days, 150 beta users joined via Threads. Giving me live insights to improve features, fix problems, and make the app better on the fly.
Learning Playground
Souko turned into a creative lab, a space where I tested AI tools, explored product thinking, and experimented hands-on. Every challenge and fix helped sharpen my skills and understanding of digital product design.
Rapid design sprint
Rapid design sprint
I kicked off Souko with a rapid design sprint powered by my own custom GPT. The goal? To design the key screens, validate my concept, and simulate early user feedback, all in record time. Instead of getting stuck in endless iterations, I used AI to generate ideas, test flows, and check accessibility right inside my sprint.
I combined this with feedback from the Threads community, turning their input into actionable insights. In less than 90 minutes, I had my first design prototype: AI-validated, user-tested, and ready to move into development. By sharing my process publicly on Threads, I tapped into the build-in-public community for fast, real-world feedback. With help from Gemini 2.0 and ChatGPT as my copilots, I was ready to jump straight into development, armed with a clear, validated concept.
(Approach Highlights)
Vibe coding
Vibe coding
As a complete beginner, I turned to ChatGPT to help me understand the best tech stack for building my first app. With the functional analysis complete, I approached development step by step, breaking it down like a game, solving small quests one function at a time.
The result? A React and JavaScript app with Firebase integration for seamless Auth0 login and database management. My goal was simple: make every piece of the app work smoothly, from saving data to handling user sessions.
(Approach Highlights)
For the first 10 days, I relied mostly on Google Gemini 2.0 in AI Studio to guide me through. It wasn’t until a few weeks in that I started using Cursor — and honestly, I wish I’d switched sooner. It boosted my workflow massively.
As I built, I gained a deeper understanding of how digital products really work. I quickly learned how to prompt LLMs effectively and get my functions running as intended. Of course, there were challenges, like spending entire nights chasing down a stubborn bug. Rookie mistakes? Definitely. But part of the process. Bit by bit, the app started to take shape. And every day, I got a little faster, a little smarter, and a lot more confident.
(Approach Highlights)
Understanding data models
Understanding data models
Design isn’t just about what users want — it’s also about what your backend can handle. Every read, every write impacts performance, cost, and user experience. Once I understood data models, I started designing smarter: thinking beyond UI, asking if features were efficient, scalable, and technically feasible. Better design = better performance.
(Approach Highlights)
From Alpha to Beta testing
From Alpha to Beta testing
Building an app tailored to your own needs is one thing. But seeing how others use it? That’s where the real learning starts.
As a designer, I love watching designs come alive in the hands of real users. Souko gave me the chance to not just build — but also test, learn, and improve. Over 150 beta testers joined the app, offering fresh insights, feedback, and ideas I hadn’t considered.
Every interaction helped me spot new opportunities, fine-tune features, and become a better designer. Honestly? That’s what makes this project so exciting for me.
(Approach Highlights)
Souko
MadeByBram © 2025