Senior Full-Stack Developer for AI Voice Coaching MVP (Next.js)

We are seeking a senior, hands-on full-stack product engineer to take ownership of an existing mobile-first sports coaching MVP and bring it to a stable pilot-ready version within three weeks. An optional fourth week may be used for stabilization and real-user feedback. This is not a greenfield project. A working Next.js prototype and initial AI functionality already exist. The selected engineer will first audit the current codebase, identify technical risks, and then improve and complete the core user experience. The Phase 1 product is intentionally focused on a simple voice-first coaching workflow. Video analysis and wearable/device integrations are outside the initial scope. Core user workflow: 1. A coach creates or selects a player. 2. The coach records natural voice feedback after a training session. 3. The application reliably captures and transcribes the audio on an iPhone. 4. AI converts the coach’s observations into a structured development priority, proposed metric, and training plan. 5. The coach can approve, edit, replace, or reject the proposed recommendation and metric. 6. The decision and session history are saved. 7. A subsequent session can be recorded to measure progress. 8. Basic structured product feedback is captured from the coach. Key responsibilities: * Audit and stabilize the existing Next.js/React/TypeScript codebase. * Implement reliable mobile audio recording, permissions, upload, and error handling, particularly in Safari on iPhone. * Integrate transcription in English and Portuguese. * Integrate OpenAI APIs and generate validated, structured responses. * Build appropriate fallback behavior when an AI request or transcription fails. * Implement or improve authentication and persistent storage for players, sessions, transcripts, recommendations, coach corrections, and follow-up results. * Deploy and troubleshoot the application on Vercel. * Test the complete user journey on mobile and desktop. * Prevent regressions and provide clear technical documentation and handover. * Communicate progress, blockers, and decisions consistently. Required experience: * Advanced Next.js, React, and TypeScript. * Strong full-stack product-development experience. * Production experience with Vercel and serverless APIs. * Experience integrating OpenAI or comparable AI APIs. * Experience with transcription, audio capture, or media uploads. * Demonstrated understanding of mobile Safari behavior. * Experience with authentication and databases such as PostgreSQL or Supabase. * Strong debugging, testing, and production-monitoring skills. * Ability to translate product requirements into a simple, reliable user experience. We are looking for someone who personally writes and owns the code. Individual senior engineers are strongly preferred. Any agency proposal must identify the exact engineer who will perform the work. Undisclosed subcontracting is not permitted. Expected engagement: * Available to begin within approximately 48 hours of selection. * Significant availability during the next three weeks. * Brief asynchronous daily updates. * Short regular check-ins with the product owner in California. * Fixed-price engagement divided into milestones. * Potential follow-on freelance work, but this is not currently a contract-to-hire position. The initial milestone will be a paid technical audit and confirmed delivery plan. Subsequent milestones will cover voice capture and transcription, the AI recommendation and coach-correction workflow, session history and progress measurement, and final pilot stabilization. The project is confidential. Shortlisted candidates will receive additional product materials and controlled access to the codebase. All work product and intellectual property must be assigned to the client, and the code must remain in client-controlled repositories and infrastructure. When applying, please answer each question specifically: 1. What production Next.js/React/TypeScript application have you personally built? Include a live link or relevant example if available. 2. Have you implemented audio recording or uploading from mobile Safari? Describe the implementation and the principal issues you encountered. 3. What have you built using OpenAI, transcription, and structured AI outputs? 4. Describe your experience deploying and troubleshooting applications on Vercel. 5. Can you begin within 48 hours, and how many hours can you realistically commit during the next three weeks? 6. Will you personally perform all development work? 7. After receiving controlled access, how quickly can you complete a code audit and produce a milestone-based delivery plan? Generic proposals that do not answer these questions will not be considered.