SpeakPing — 30-Second Daily Speaking
A mobile app that turns speaking practice into a 30-second daily habit — one AI question, one recording, instant feedback, across 7 languages.
/ Overview
SpeakPing is a mobile app that turns speaking practice into a 30-second daily habit. Each day it generates one AI spoken question; the learner records a 30-second answer and gets instant scoring on fluency, vocabulary, grammar, and pronunciation — across 7 languages, with streaks and reminders that keep them coming back.
Get in touch/ Project Details
Name:SpeakPing — 30-Second Daily Speaking
Client:SpeakPing
Industries:EdTech / Language Learning, AI
Date:2026
Project page:https://speakping.com
/ Challenge
Most language learners can read and write but freeze the moment they have to speak. Grammar drills never build real speaking confidence, scheduling live practice is hard, and learners rarely know if they're improving. SpeakPing removes every barrier with a tiny daily loop and instant, objective AI feedback.
/ Provided Services
30-Second Daily Loop
One AI-generated spoken question, one 30-second recording — a habit small enough to keep every single day.
AI Feedback in Seconds
Instant scoring on Fluency, Vocabulary, Grammar, and Pronunciation so learners know exactly what to improve.
7 Languages, 1 App
Practice English, Vietnamese, Spanish, German, Japanese, Korean, or Chinese in one beautifully simple app.
Streak & Gamification
Daily streaks, smart reminders, and progress tracking that turn practice into a sticky habit.
.webp&w=3840&q=75)
.webp&w=1920&q=75)
.webp&w=1920&q=75)
/ Solution
A cross-platform React Native app backed by a NestJS AI pipeline delivers real-time speech scoring at scale. On-device Whisper handles speech-to-text, while a NestJS + TypeORM backend orchestrates OpenAI and Google Gemini for question generation and feedback, the Microsoft Speech SDK for pronunciation assessment, Qdrant for retrieval, and Firebase for auth, messaging, and analytics.
/ Tech stack
React Native (Expo)
Cross-platform mobile app with React Query, Firebase auth (Google/Apple), and expo-av audio recording.
NestJS + TypeORM
Structured backend API for questions, scoring, streaks, and subscriptions.
OpenAI + Google Gemini
AI question generation and multi-dimensional speaking feedback.
Microsoft Speech SDK + Whisper
Pronunciation assessment plus on-device speech-to-text.
Qdrant + AWS S3
Vector retrieval and audio storage.
Firebase
Authentication, push messaging, and analytics.
