Contact person: AI-Powered Speaking Practice App

Phone:Show

Email:Show

Location: Hong Kong

Budget: Recommended by industry experts

Time to start: As soon as possible

Project description:
"I want to create a mobile language-learning app whose single, clear purpose is to help users practise speaking. The core experience should feel like having a private coach: a learner speaks into the phone, our AI engine listens, evaluates, and then returns actionable feedback in real time.

What matters most is an accurate, friendly assessment of two areas: pronunciation accuracy and overall conversational fluency. I will rely on the app to record a short snippet, run voice-to-text, compare the result to the target phrase or prompt, score the user, and present concise suggestions—think mispronounced phonemes highlighted and a fluency meter that tracks pace, pauses, and filler words. A short audio playback of the user’s own recording, paired with the ideal model pronunciation, would be ideal so users can immediately hear the gap.

Although the radio question mentioned voice recognition, live tutors, and pre-recorded dialogues, I specifically chose an AI-driven assessment instead. That means the priority is integrating a robust speech-to-text engine (Google Speech, Azure, or DeepSpeech are all fine as long as they handle multiple accents) and an NLP layer that can parse, score, and display results quickly on both iOS and Android.

Deliverables will be straightforward:
• A working cross-platform prototype (React Native or Flutter preferred, but I’m flexible) that covers user login, a simple dashboard of speaking exercises, live recording, AI scoring, and feedback screens.
• Clean, well-commented source code in a private Git repo.
• A short technical read-me that documents model selection, scoring rubric, and how to add new prompts in the future.

When the prototype feels solid—low latency, reliable scoring, clear feedback—we can discuss polishing the UI and expanding to listening comprehension, but for now let’s keep the scope tightly focused on speaking practice powered by AI." (client-provided description)