Back to Journey
December 2025MILESTONE

Building My First AI-Powered Web App

From AI curiosity to shipping a healthcare solution

The Project: GlucoCheck

Built GlucoCheck to learn how to integrate AI into web applications. Created a diabetes risk assessment tool where users fill out a health questionnaire, and the AI analyzes their responses to generate personalized risk assessments in Bahasa Indonesia. Learned how to work with AI APIs, prompt engineering to get clear and helpful responses, and present complex AI outputs in a user-friendly way. First real experience wrapping AI capabilities into an actual product.


How It Works

Users answer a series of health-related questions, and the backend sends this data to an AI service (like OpenAI’s API) with a carefully crafted prompt. The AI analyzes the responses considering factors like age, BMI, family history, physical activity, and diet. It then generates a personalized risk assessment with recommendations. The tricky part wasn’t the frontend or backend — it was the prompt engineering. Getting the AI to return responses that were accurate, helpful, culturally appropriate (in Bahasa Indonesia), and formatted in a way the frontend could display cleanly.


Technologies & Skills


Key Insights

  • Prompt Engineering is Crucial: The quality of AI output depends heavily on how you phrase the prompt
  • User Experience Matters: Presenting AI outputs in a way users can understand and act on is equally important as the AI itself
  • Localization is Important: Generating outputs in Bahasa Indonesia made the app accessible to a wider user base
  • AI is a Tool, Not Magic: It’s powerful but requires thoughtful integration and validation

Impact & Takeaway

GlucoCheck taught me that AI isn’t just about fancy algorithms — it’s about solving real problems for real people. Building a health assessment tool meant understanding that my code could actually impact someone’s health decisions, so accuracy and clarity were critical. This project sparked my obsession with AI, making me realize that the most interesting problems to solve are those that combine strong product thinking with powerful AI capabilities. It also made me hungry to go deeper into ML and AI engineering, which led me to apply for the DBS Coding Camp.

Part of my development journey