Logtree: From Zero to App Store
in Six Weeks
A case study in modern AI-assisted development
It is a common refrain in the product management world. To truly understand the engineers, you must understand the code. Cong, a seasoned Product Leader based in Singapore with 16 years of experience, took this advice to its logical extreme. He didn't just read a book on Swift. He built and shipped an iOS app from scratch, with zero prior coding experience.
The app is called Logtree. It is a delightful, retro-styled logging utility that lets you track anything, including habits, fasting, and data trends. It is wrapped in a UI that pays homage to the legendary DOS file manager, XTree Gold.
The interesting part of this story isn't just the app itself. It is how it was built. This is a case study in the modern "AI-assisted" development workflow.
Cong's journey began where many non-technical founders start: the "No-Code" ecosystem. He initially reached for FlutterFlow, a drag-and-drop builder, hoping to bypass the syntax errors and compiler warnings of traditional development.
It worked, until it didn't.
The limitations of the sandbox became apparent quickly. As soon as he wanted to implement complex data visualizations, the kind that makes a logging app actually useful, he hit a wall. As Cong put it:
"FlutterFlow seemed great at building template apps, but moving outside of that sandbox required custom code... Giving up on no-code made me feel like I had an ocean to cross on a rowboat."
This is the classic trap of abstraction layers. They are wonderful for prototypes. However, when you want craftsmanship and custom behavior, you eventually have to write the code.
So, he pivoted. He decided to learn Swift and SwiftUI properly. Unlike the developers of a decade ago who relied solely on Stack Overflow and brute persistence, Cong enlisted a team of AI assistants: ChatGPT, Claude, and eventually the Cursor IDE.
The workflow he describes is fascinating. He started with ChatGPT (specifically the o1-preview model) to generate pseudocode and explain concepts. He then moved to Cursor, an AI-first code editor that integrates LLMs directly into the codebase. This allowed him to stop pasting snippets back and forth. Instead, he could let the AI analyze the entire project context.
It wasn't magic, though. The LLMs, as impressive as they are, are prone to "hallucinations" where they confidently spit out code that compiles but doesn't work. When transitioning from Firebase to Core Data for offline-first support, the AI led him down several blind alleys.
This forced Cong to actually learn. He couldn't just copy-paste. He had to watch tutorials (shout out to iOS Academy) to understand the foundational concepts of data modeling. Only then could he prompt the AI correctly to get the right answers.
His advice to aspiring AI-assisted developers is spot on: Type the code out manually. Don't just command-C, command-V. Typing it forces you to read it, line by line, and understand the logic flow.
Regarding the design, Cong eschewed the standard "Clean functionality" of modern iOS guidelines. Let's be honest, those can feel a bit sterile. He opted instead for a brutalist, retro aesthetic.
He modeled the UI after XTree Gold, the 1990s file management utility. It features a folder tree view, stark contrasts, and a "function over form" vibe that ironically requires more form to implement today. Because standard UIKit and SwiftUI components are designed to look like modern iOS, Cong had to rebuild many UI components from scratch to achieve that authentic retro feel.
In just six weeks, working 2-4 hours a night after putting the kids to bed, Cong went from "Hello World" to a live App Store listing. He estimates the process would have taken six months without LLM assistance.
It is a testament to the power of these new tools. They don't replace the need for domain knowledge. You still need to know what to ask. But they flatten the learning curve from a steep cliff to a "hockey stick."
For a Product Manager, the exercise was invaluable. "Understanding the complexities [engineers] face makes you a better collaborator," Cong says.
Check out the LogTree site for more information about the app, or just download it straight from the App Store today!
Download