Projects

AI Prototype Generator

Role: Full-Stack Developer & UX Designer | [Demo Video] [GitHub] [Retrospective]

One-Line Summary: An internal discovery tool that leverages generative AI to instantly translate abstract user problems into functional action items.

What it delivers: Accelerates the sales and scoping cycle by instantly translating a client's business challenge into a working technical proof-of-concept. By eliminating days of cross-functional alignment, it slashes time-to-value and streamlines team collaboration.

Architecture:

  • Frontend: React interface featuring real-time component hierarchy visualization and live code previews.
  • Backend: FastAPI (Python) service to manage asynchronous generation queues and prompt orchestration.
  • AI Integration: Gemini API using strict structured JSON schemas to enforce syntactically consistent UI rendering.

Impact & DX: Built a zero-config, friction-free developer experience that cut prototyping onboarding to under a minute. Implemented strict LLM exception handling and parsing fallbacks to secure a 99% successful component rendering rate.

Footnotes (Spotify API Visualizer)

Role: Developer | [GitHub]

One-Line Summary: A dynamic data-visualization application that processes raw streaming metadata into an interactive timeline mapping a user's evolving musical tastes.

What this delivers: Displays listening shifts and underlying behavioral trends, cementing the memories of month.

Architecture:

  • Authentication: Spotify OAuth 2.0 protocol for scoped user authentication and token refresh flows.
  • Data Pipeline: Developed logic to extract, filter, and map deeply nested JSON payloads from Spotify’s personalization endpoints.
  • UI/UX: Designed a minimalist, accessible dark-mode UI optimized for and fast client-side data ingestion.

Impact & Communication: Successfully bridged the technical-to-user gap by translating abstract machine-readable metrics (e.g., "0.82 valence") into visual stories.

Explore more projects on GitHub →

Home