Full-Stack Developer for AI-Powered Marketplace Platform

We're building an AI-powered marketplace platform and are looking for an experienced full-stack developer to lead development from MVP through launch. The platform connects buyers and sellers, with AI features such as intelligent matching, automated listing generation, conversational search, and recommendation/ranking. Tech stack (required experience): - Frontend: Next.js (App Router), React, Tailwind CSS - Backend/API: Node.js and Python (FastAPI or similar) - Database/Auth/Storage: Supabase (Postgres, Row Level Security, Edge Functions, Realtime) - AI: OpenAI and/or Anthropic APIs — prompt design, structured outputs, tool/function calling, embeddings and vector search (pgvector), RAG pipelines - Payments: Stripe (Connect for multi-vendor payouts preferred) - Deployment: Vercel, plus a Python service host (Railway, Fly.io, or AWS) Scope of work: 1. Set up project architecture, repo, CI/CD, and environments 2. Build core marketplace features: user roles (buyer/seller/admin), onboarding, listings, search and filters, messaging, orders/transactions, reviews, admin dashboard 3. Implement AI features: semantic search over listings, AI-assisted listing creation, matching/recommendations, and an in-app assistant 4. Integrate Stripe for payments and seller payouts 5. Write clean, documented, testable code; set up basic test coverage 6. Handle security basics: RLS policies, input validation, rate limiting, secret management Must-have qualifications: - 4+ years full-stack experience, with at least 2 shipped production apps using Next.js + Supabase - Hands-on experience building features on top of OpenAI or Anthropic APIs (please link to a live example or repo) - Strong Postgres/SQL skills - Comfortable owning a project end to end with minimal supervision - Clear written English and reliable async communication Nice to have: - Prior marketplace or two-sided platform experience - LangChain/LlamaIndex, vector databases, or evaluation tooling for LLM features - TypeScript throughout, Zod/tRPC, Playwright testing - UI/UX sensibility (shadcn/ui, Figma handoff) To apply, please include: 1. Links to 1–2 projects built with Next.js + Supabase (live if possible) 2. One example of an AI/LLM feature you've shipped and how you handled prompt design, cost, and reliability 3. Your proposed high-level architecture for this stack in 3–5 sentences 4. Your availability and rate 5. Start your proposal with the word "Marketplace" so we know you read the full post We'll shortlist quickly and do a short paid trial task before a longer engagement.