Full-Stack Developer — AI Chat Platform (Next.js + Supabase + OpenAI)

I’m building a structured productivity and coaching platform centered around an AI-powered chat experience. The AI guides users through goal-setting, reflection, and progress tracking, with full conversation history, context persistence, and personalized program content stored and retrieved using PostgreSQL with pgvector. This is a real product built to scale, not a prototype. The stack: Next.js (App Router), Node.js, PostgreSQL, pgvector, Supabase, OpenAI API (streaming completions), Clerk or Auth.js for authentication, and Vercel for deployment. What the role requires: Expert-level Next.js full-stack development, solid PostgreSQL schema design, hands-on OpenAI API integration with streaming and conversation context management, and Supabase experience including row-level security. Familiarity with RAG patterns using pgvector is a strong plus. How I work: I am not a developer, but I am actively involved in architecture, design decisions, and QA. I will not hand this project over and hope for the best. I need a developer who communicates proactively, explains their decisions in plain language, flags issues early, and treats this as a collaboration. If you prefer to work in isolation and surface results at the end, this isn’t the right fit. To apply: Skip the generic proposal. Tell me about an AI chat or streaming product you’ve built — the stack, the challenges, and what you’d do differently. Give me your honest take on the stack I’ve described — do you agree with these choices, or would you change anything? Applications that don’t address these questions won’t be considered.