TRADIX

• 

  Mockup TRADIX app

Inspiration In the dynamic Indonesian stock market, retail investors often find themselves at a disadvantage. They face a market with unique challenges, like the influence of "Bandar" (market manipulators), and a lack of access to the sophisticated analytical tools available to professional traders. We were inspired to level the playing field. Our vision was to create TRADIX, an all-in-one, AI-powered investment platform that not only provides powerful tools but also educates users, empowering them to make smarter investment decisions with confidence.

How We Built It We built TRADIX from the ground up as a cross-platform mobile application using React Native and Expo, ensuring it's accessible to everyone. Our development was an intense and iterative journey focused on integrating cutting-edge features:

AI-Powered Analysis: We integrated Google's Gemini AI to provide users with intelligent trading signals, a smart stock screener, and deep chart pattern analysis. The "Bandar Detector": A unique feature tailored for the local market that analyzes market flow to detect the activity of large players, giving retail investors unprecedented insight. Advanced Tools & Education: We coupled these AI features with professional-grade charting, real-time data, and a comprehensive education module to create a holistic investment ecosystem. The process involved rapid prototyping, rigorous testing, and constant refinement. We went from simple data fetching from sources like Yahoo Finance to building a robust, real-time data pipeline.

What We Learned This project was a tremendous learning experience in building a complex, data-centric application.

The AI Balancing Act: We initially experimented with different AI providers but learned the critical importance of balancing performance, cost, and analytical quality. This led us to switch to Gemini AI and implement a sophisticated caching strategy to optimize API calls, making powerful AI features financially viable. Data is King, and It's Ruthless: Building a financial application taught us that data integrity is non-negotiable. We spent a significant amount of time fixing subtle bugs, like NaN errors in financial calculations and ensuring the accuracy of real-time data, which reinforced the need for meticulous data validation. Architecture Evolves: A project's architecture isn't set in stone. The decision to remove a specific AI provider and build our own caching layer was a pivotal moment that significantly improved our application's performance and scalability. Challenges We Faced Taming the Data Dragon: Ensuring the accuracy, timeliness, and reliability of financial data from multiple sources was our biggest and most persistent challenge. Decoding the "Bandar": Creating a feature to analyze and visualize the complex patterns of market manipulators was a unique data science challenge that required deep domain knowledge and innovative analytical techniques. Performance Under Pressure: Delivering a smooth user experience in a mobile app that is constantly processing real-time data streams and running complex AI analyses required continuous performance audits and optimizations.

Built With

• android
• claude
• cloudflare
• expo.io
• ios
• javascript
• jest
• reactnative
• supabase
• tensorflow
• typescript
• yahoofinanceapi