Senior Full-Stack Engineer — AI-Native Real Estate SaaS (Next.js / TypeScript / Supabase)

About the role We're building an AI-native SaaS platform for the real estate industry — an intelligence layer that helps agents run their entire business. The product is deeply AI-driven: large language models power real, production-facing features, not bolt-on gimmicks. We're in the final stretch — MVP is weeks from launch — and I'm building a long-term team to get it across the finish line and keep building and supporting it well beyond launch. This isn't a one-off gig. I want people who want to own their work, grow with the platform, and still be here as we scale post-launch. You'll build production features across a modern stack, working against clear specs with code review on every PR. This is execution-focused work — you own well-defined modules and ship them clean. You don't need to architect our AI system, but you should be comfortable working around LLM-powered features: calling APIs, handling structured responses, and building the surfaces on top of them. The stack Next.js (App Router) · TypeScript · Tailwind Supabase (Postgres + RLS + Realtime) · Vercel LLM integration via API (Anthropic / similar) · third-party integrations (payments, comms, calendar) What you'll do Build production full-stack features against detailed specs Integrate and surface AI-powered functionality (API calls, structured output handling, UI on top of model responses) Wire up third-party APIs cleanly Write tested, maintainable code that passes review Communicate clearly and hit timelines Requirements 7+ years full-stack, strong in Next.js + TypeScript Hands-on Supabase or PostgreSQL (RLS a plus) Experience integrating LLM / AI APIs into production apps Public GitHub with real, reviewable code Meaningful daytime overlap with US Eastern time Fluent written English, responsive communication To apply — read carefully (proposals that skip this are ignored): Start your proposal with the word "Compass" so I know you read this. Link your GitHub and point me to one project you built — tell me what you specifically did. One sentence: how you'd handle an LLM API returning malformed JSON in production. We review every application that follows all three steps.