Building intelligent systems that make a difference.
Portfolio
Blog
TTS implementation and real-device testing. Web Speech API platform differences, mobile browser constraints, and the path to a native app.
Testing local LLM on RTX 3060, mobile LLM investigation, and future plans including TTS, bidirectional translation, and smart glasses.
Technical architecture behind Intent-First Translation: streaming LLM output optimization, 3-layer design, and 6 model benchmarks.
Why existing voice translation apps fail at natural conversation, and how Intent-First Translation approaches the silence problem.
