Wordapp
An AI-powered word of the day web app, powered by Gemini AI.
Technologies Used
TypeScriptPythonAstroTailwind CSSStarwind UIGemini AIMongoDBFlexSearchZodPydanticGitHub Actions
Timeline
Aug 2026 - Aug 2026Co-Authors
This project was completed individuallyNo video available
Key Accomplishments
- Developed a Python-based word curation pipeline that reduced an initial corpus of over 235,000 words to roughly 500 high-quality vocabulary candidates using lexical, semantic, and frequency-based filtering.
- Optimized the curation pipeline from roughly 200-240 seconds to around 3 seconds by caching expensive lemmatization, frequency, and Open English WordNet lookups.
- Developed a Gemini-powered generation pipeline to produce structured vocabulary metadata with schema validation and reusable generation workflows.
- Automated daily word generation and MongoDB persistence with GitHub Actions, including scheduled generation and support for generating words for any provided date.
- Integrated FlexSearch with a pre-build indexing step to provide fast client-side search across generated vocabulary data.
- Built the frontend with Astro, Tailwind CSS, and Starwind UI, including daily word, archive, glossary, word detail pages, text-to-speech pronounciation of words, RSS feeds, and SEO functionality.
- Leveraged spec-driven development with OpenSpec to define feature requirements, acceptance criteria, and implementation tasks before development, providing traceability from specification through implementation.