Speak Core Workflow B: Pronunciation Training

Overview

Secondary workflow for Speak: detailed pronunciation training with phoneme-level analysis and adaptive practice. Uses OpenAI's speech recognition with Speak's proprietary proficiency graph to identify and drill weak phonemes.

Prerequisites

• Completed speak-core-workflow-a
• Audio recording capability (WAV 16kHz mono)
• ffmpeg installed for audio preprocessing

Instructions

Step 1: Pronunciation Assessment

import { SpeakClient } from '@speak/language-sdk';