How I automated IPA transcription for linguistics: A story of CMUdict and Offline-first design

As anyone in the linguistics or ESL (English as a Second Language) field knows, manually transcribing full paragraphs into the International Phonetic Alphabet (IPA) is a tedious, error-prone nightmare. Most online tools are designed for single-word lookups. But what happens when a teacher needs to prepare a 5-page handout with word-aligned transcriptions? They usually end up fighting with Word fonts, broken Unicode symbols, and chaotic alignment. I decided to fix this by building Phonetic Formatter . The Challenge: Accuracy vs. Privacy The first hurdle was the data source. While many look toward cloud APIs, I wanted this to be 100% offline . Teachers and students often work in environments where privacy and connectivity are concerns. I chose the CMU Pronouncing Dictionary as the foundation. It’s an incredible open-source resource, but mapping it to a high-performance mobile interface required careful optimization to ensure near-instant batch conversion of long texts without draining th