https://github.com/asmuelle/parlo

An AI speaking partner for language learners that runs entirely on the phone — unlimited conversation roleplay and pronunciation feedback, offline, at a quarter of Speak's price.
https://github.com/asmuelle/parlo

android edge-ai foundation-models ios language-learning on-device-inference speech-recognition

Last synced: about 18 hours ago
JSON representation

An AI speaking partner for language learners that runs entirely on the phone — unlimited conversation roleplay and pronunciation feedback, offline, at a quarter of Speak's price.

• Host: GitHub
• URL: https://github.com/asmuelle/parlo
• Owner: asmuelle
• Created: 2026-06-11T16:50:55.000Z (26 days ago)
• Default Branch: main
• Last Pushed: 2026-06-19T23:17:05.000Z (18 days ago)
• Last Synced: 2026-06-20T01:10:21.537Z (18 days ago)
• Topics: android, edge-ai, foundation-models, ios, language-learning, on-device-inference, speech-recognition
• Language: Swift
• Homepage: https://asmuelle.github.io/parlo/
• Size: 443 KB
• Stars: 0
• Watchers: 0
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • Agents: AGENTS.md

Awesome Lists containing this project

README

          
# Parlo

[![CI](https://github.com/asmuelle/parlo/actions/workflows/ci.yml/badge.svg)](https://github.com/asmuelle/parlo/actions/workflows/ci.yml)

> An AI speaking partner for language learners that runs entirely on the phone — truly unlimited conversation roleplay and pronunciation feedback, offline on the commute or abroad, at a quarter of Speak's price.

**Category:** Edge AI / on-device inference (iOS + Android) 

## Concept

An AI speaking partner for language learners that runs entirely on the phone — truly unlimited conversation roleplay and pronunciation feedback, offline on the commute or abroad, at a quarter of Speak's price.

## Target User

Intermediate (B1-B2) learners who need speaking volume above all: travelers, expats, and immigrants currently paying Speak $20/mo or Duolingo Max $30/mo and rationing metered practice minutes — plus shy learners who won't stumble through bad sentences into a cloud microphone.

## Why Edge AI Is Structural (not decoration)

Whisper-class on-device ASR with phoneme-level forced alignment for pronunciation scoring; AFM 3 Core / Core Advanced for conversation roleplay via Dynamic Profiles (café scene, job interview, haggling) with guided generation for inline corrections; Gemma 3n E2B audio-input multimodality via LiteRT-LM on Android lets learners speak directly to the model; on-device TTS replies. Structural: speaking practice is the highest-token activity in language learning — incumbents meter minutes precisely because every cloud turn costs money, so 'truly unlimited speaking' is a feature cloud economics forbid; and learners practice exactly where connectivity dies (planes, subways, abroad without data).

## Why Now (2026 timing)

Gemma 3n audio input shipped and Gemma 4 (140+ languages) is in AICore Developer Preview; A19 Pro makes 20B-class on-device conversation quality real; subscription fatigue is quantified ($66/mo average AI spend, 53% cancel-and-restart), making a cheap annual/lifetime tutor the structural answer.

## Tech Stack

iOS-first (primary platform, zero inference cost): Swift/SwiftUI; FoundationModels framework (iOS 26+, ~3B AFM) with @Generable guided generation emitting structured turns {reply, correction, naturalness note}; aggressive transcript condensation per Apple TN3193 to live inside the 4096-token context window; SpeechAnalyzer/SpeechTranscriber (iOS 26 on-device ASR) for supported locales with whisper.cpp small/medium (Metal) fallback for others; pronunciation scoring via a Core ML-converted wav2vec2 forced-alignment + GOP model, shipped explicitly as beta-grade syllable feedback, not Speak-parity scoring; TTS via AVSpeechSynthesizer premium voices or Kokoro-82M through MLX/Core ML. Android (phase 2, flagship-gated): MediaPipe LLM Inference API / LiteRT-LM running Gemma 3n E4B on GPU/NPU (hard-gate on 8GB+ RAM devices, Play Asset Delivery for the 3-4GB weights), Gemma 3n audio input or whisper.cpp for ASR, Android on-device TextToSpeech; adopt Gemini Nano/ML Kit GenAI and Gemma 4 via AICore only where the allowlist permits. Pragmatic escape hatch that admittedly dilutes the pure-edge story: optional Gemini Flash cloud path for unsupported devices/languages, off by default and clearly labeled. Launch scope: 2-3 high-resource target languages (Spanish, French, German) where 3B output quality is least embarrassing; expand only after per-language native-speaker QA.