An open-source, real-time streaming Automatic Speech Recognition (ASR) model for Thai, optimized for low-latency CPU deployment.

A Next.js and FastAPI demo application for Typhoon ASR - Thai speech recognition in real-time.

This project was developed during my internship. It focuses on transcribing Thai stock analysis audio to text using Typhoon ASR, incorporating Gemini Flash for verification, fact-checking, and summarization.

An open-source, real-time streaming Automatic Speech Recognition (ASR) model for Thai, optimized for low-latency CPU deployment.

A benchmarking script to determine the optimal configuration to achieve the highest RTFx for Typhoon-ASR model