1. Overview
FINGOO is a service that brings investment learning and AI-powered investment analysis together in a single app.
While studying at the University of Seoul, I was approached by the team's marketer, who had noticed my activity on GitHub, and invited me to join the project.
I'd always been interested in investing, so the topic itself was compelling, and I saw it as a chance to grow as a developer by working on a real, live product — so I decided to join.
I came on as a frontend developer in May 2026. Version 3 is currently live, and we're now building version 4, targeting a release in early 2027.
2. System Architecture
I'm responsible for frontend development on the team, and the app is built on React Native and Expo. Expo has recently improved so much as a developer experience that it's even recommended in the official React Native docs, which made it feel like a safe choice.
For styling, I introduced Nativewind, which offers syntax similar to Tailwind CSS, to keep our design consistent across the team as we collaborated.
For client-side state management, we decided to use Zustand, a widely adopted library. We also introduced Tanstack Query for server state management to handle caching and synchronization efficiently, and set up a test environment with Jest.
FINGOO is made up of the following screens.
| Screen | Role |
|---|---|
| Home screen | Ask the conversational AI for investment analysis, with suggested questions |
| Learning screen | Learn investing fundamentals step by step through card-based content |
| Market analysis screen | Get real-time data and charts in response to questions about specific stocks |
| Ranking screen | Earn XP through attendance checks and quizzes, and compete on the leaderboard |
3. Retrospective
I assumed Nativewind would behave exactly like Tailwind CSS, but while developing I found that the two libraries actually have different default scale values in places. I shared this with the team, which helped us avoid style inconsistencies down the line.
Working as a frontend developer on a product with real users, from early on, gave me a different sense of responsibility — and different lessons — than my previous personal projects.
Using the feedback we've gathered from running version 3, we're now building a better version 4.