Speech Model API + react native integration for mobile app

Title: ML/Backend Engineer — Speech Model API + React Native Integration (Arabic Pronunciation App) Description: I'm building a mobile app that gives users real-time feedback on spoken Arabic pronunciation, using a custom-trained speech recognition model. Tech stack (required): Backend/ML serving: Python, FastAPI (or equivalent) Mobile app: React Native (iOS + Android) Please only apply if you have real, demonstrable experience with React Native specifically — not Flutter, not native-only iOS/Android, Model & API : I have an existing fine-tuned wav2vec2-based model (trained on a substantial labeled dataset of correct/incorrect recitations) that needs refinement — known issues include some accuracy bias on certain speech patterns, and improving classification on closely related/confusable sounds Needs a production-ready API (FastAPI or your recommended equivalent), with attention to real-time/near-real-time latency for a live, repeated user interaction Confidence-based output classification (multiple feedback tiers, not just binary correct/incorrect) — full detail shared under NDA Clear, documented API contract for downstream mobile integration Mobile integration : React Native mobile app calling the above API, built against a finalized Figma design (design system, all screens, and states already specified) Standard auth, user progress tracking, subscription/paywall What I'm looking for: Real experience fine-tuning or self-hosting a speech/ASR model (not just integrating third-party APIs like Whisper, Deepgram, or AssemblyAI) PyTorch/HuggingFace model serving experience If applying for or interested in Phase 2, genuine React Native experience is required — please share app(s) you've shipped in React Native specifically Bonus: experience with non-standard/accented speech populations Please include in your application: A specific example of a self-hosted speech/ASR model you've fine-tuned or deployed (not a third-party API integration)