Full-Stack Support Engineer — Part-Time / On-Call

We run a small, profitable, production SaaS platform used every day by property operators to manage their portfolios — messaging, billing, access control, financial reporting, and a handful of partner integrations. It's a real product with paying customers who depend on it, which means when something breaks, it needs a calm, fast hand. We're looking for one experienced full-stack engineer to own bug triage, integration health, and the "first responder" work a live multi-tenant product needs. This is not 24/7 pager duty — think "responds thoughtfully within a few hours, most days," on a part-time retainer. You'd work directly with the founder. HOW WE WORK We build and maintain this product AI-first. You'll lean heavily on agentic coding tools and Replit — where our dev and deployment both live. We care far more that you're a genuine AI power user than that you can recite language trivia. That said, this is a debugging role at its core — AI fluency layered on top of real skill, not in place of it. MUST-HAVES - AI power user (Cursor, Claude Code, Codex) — weighted above raw language proficiency - Experienced Replit user — you know the environment and its quirks - TypeScript — full-stack, no JavaScript-only candidates - Python — background jobs and browser-automation tooling - React — hooks, TanStack Query (v5), Radix/shadcn - Node.js + Express — REST routes, middleware, session-based auth - PostgreSQL — real queries, indexes, query plans - Vitest — you write and fix tests, not just run them - Integration debugging — webhooks, HMAC, API tracing - Background jobs & scheduled tasks — cron, queues, retry logic, timezone bugs - AWS CloudWatch / serverless logs — Lambda-style diagnosis YOU'LL BE DEBUGGING THINGS LIKE - Stripe Connect: webhook delivery, payout splits, proration, billing cycles - SMS/email delivery failures and provider rate limits - Smart-lock token refresh and webhook record delivery - Partner & data API sync, geocoding, AI-generated content timeouts - Banking webhooks — HMAC verification and expense materialization - Serverless delivery functions and background workers THE RIGHT LEVEL Mid-to-senior. Someone who's shipped and maintained production software and actually enjoys being the person who finds the root cause. SCHEDULE Monday–Friday, US Central Time, with weekends as needed. Meaningful CT overlap required regardless of where you're based. HOW TO APPLY Send a short note and answer this question — it's the real filter: "Describe a time you debugged a webhook that was silently failing. What tools did you use, how did you find the root cause, and what did you do to keep it from happening again?" Also tell us: how do you actually use AI in your day-to-day engineering? Specific, concrete answers move forward. Generic ones don't.