Pronunciation Feedback System Development

Edwin King'ori

Backend Engineer

Docker

Docker Compose

FastAPI

Electronics

The Goal

To build a backend-powered pronunciation feedback system that analyzes a user’s spoken English audio, compares it to the expected text, and returns detailed phoneme-level feedback.

Sample Request

{ "text": "Hello there! It's nice to finally meet you.", "audio": "<base64-encoded WAV>" }

Expected Response

{ "phrase": "Hello there...", "phonemes": [ {"phoneme": "h", "start": 0.12, "end": 0.18, "status": "correct"}, {"expected": ["e", "l"], "actual": ["æ"], "status": "wrong"}, {"expected": ["oʊ"], "status": "missing"}, ... ] }

Montreal Forced Aligner

https://montreal-forced-aligner.readthedocs.io/en/latest/installation.html#docker-installation

Phonemizer

https://github.com/bootphon/phonemizer

Like this project

Posted Oct 6, 2025

Built a backend system for phoneme-level pronunciation feedback.

Likes

Views

Timeline

Jul 8, 2025 - Jul 15, 2025