Project Description
I’m building a real-time speech-to-text application for Tamil and need a full mobile solution that runs smoothly on both Android and iOS. The core requirement is low-latency live transcription that recognises the major dialects of Tamil—Madurai, Kongu, Nellai, Chennai and Sri Lankan variations—so users hear their words appear on-screen almost instantly, regardless of accent.
My priority is accuracy and speed, followed by an interface that keeps the mic open, shows streaming text, and lets users copy, save or share the transcript once they stop speaking. If you can add useful extras such as offline mode, punctuation handling, or a light / dark theme switcher, feel free to mention them.
When you respond, focus on your relevant experience: the speech-to-text engines you’ve integrated (Google Speech, Azure Cognitive, Mozilla DeepSpeech, etc.), any work you’ve done with Indian languages, and how you would approach dialect modelling and latency optimisation in a cross-platform codebase (Flutter, React Native, or native twins—your call).
Deliverables should include:
• Source code ready to compile for Android and iOS
• A brief setup guide and model/engine configuration
• A demo build or TestFlight link showing real-time Tamil transcription across at least three dialects
I look forward to seeing how your experience can bring this idea to life.