My personal projects are not just hobbies. They are practical experiments for understanding how new technologies can be applied in real product and business contexts.

Project: SUNDAY MUSIC COWNTDOWN

URL: https://sundaymusic.yre.jp/

Overview This is a PoC project that fully automates the entire content operation workflow, from “collection, selection, summarization, to distribution of information,” using Generative AI (Gemini API) and AWS serverless services. It builds a pipeline where AI structures unstructured data (music trends), reconstructs it into user-value content, and continuously delivers it.

Architecture & Tech Stack

• Core Logic: Python (FastAPI)
• AI / LLM: Google Gemini API (Summarization, Tagging, Metadata Generation)
• Compute: AWS Lambda (Serverless Batch)
• Database: Amazon DynamoDB (NoSQL)
• Hosting: Amazon S3 + CloudFront
• Key Achievement: Realized a microservices architecture that keeps operating costs within the AWS Free Tier through a serverless design.

Project: What to Do with Inherited Land?

URL: https://souzoku-tochi.jp

Overview An information site for people struggling with how to dispose of inherited land, providing institutional comparisons, diagnostics, cost simulations, and vendor information. It translates a highly specialized domain into practical, easy-to-understand content and is continuously updated.

Architecture & Tech Stack

• Frontend Framework: Next.js 14 (App Router) + React 18 + TypeScript

• Styling: Tailwind CSS (+ PostCSS / Autoprefixer)

• Content Management: Type-safe article data managed in src/data/blog.ts (Markdown-like body content managed in code)

• SEO / Discoverability: Dynamic sitemap.ts, robots.ts, and article metadata design

• Analytics: Google Analytics (gtag.js)

• Deployment / Hosting: Vercel

• Key Achievement: Achieved both type-safe content operations and SEO-oriented design without depending on a CMS. Built a lightweight, maintainable platform for fast publishing and updates of specialized articles.