Senior FullStack / Platform Engineer for Solo Research Tech Founder Launch + Reliability (US/LatAm)

I'm the founder of a solo-operated research operations platform company. The platform is in production and serving clients. Much of it was built fast with AI coding agents, and it now needs a senior engineer to do a rigorous technical pass, get it launch-ready for a high-stakes client event at the end of June, and then own its ongoing health. I need someone who can hold the whole picture: not just the infrastructure, but the application code and the data flow end to end. You'll work alongside me, an Operations Coordinator, and a small set of AI coding agents that handle a lot of the build. I need someone senior enough to own technical decisions without hand-holding, and disciplined enough to work safely in a codebase where a careless change has real consequences (more on that below). The near-term mission (June): I have a national client running a live, in-person conference at the end of June where a family of 8 surveys must go live with real participants and feed a reporting dashboard correctly. Your first job is to make that happen and make it boring: - Do a full technical pass across the monorepo and certify launch readiness: surveys render and submit cleanly, data flows survey to analytics to dashboard without loss, and the known gaps are closed. - Hunt and fix the application-level bugs that block a clean launch (renderer edge cases, unwired hooks, integration regressions, data-integrity gaps in the multi-survey campaign flow). This is real full-stack debugging in Next.js, React, and TypeScript, not just infra work. - Lead the technical integration of our client-facing site (built on Vercel) into the client's own environment: DNS, redirects, embedding, and auth, confirmed against their requirements before the event. You'll pair with my Operations Coordinator and liaise directly with the client's technical contact to get this right. I'm on the key calls with you. - Run the end-to-end QA: take the surveys as a participant across locales and cohorts, confirm the dashboard reflects reality, and sign off that it's issue-free. What you'll own at steady state: - Platform infrastructure across a multi-app monorepo (Vercel, Railway, Next.js, Postgres, Prisma, Clerk). - Application reliability, not just uptime: the code paths that move data from survey to dashboard have to stay correct as the platform grows. - A regular monthly health pass that keeps the platform shippable. Incident response on production, with a 4-hour response target during business hours. - DevOps hygiene: environment config, secrets and integration token rotation, deploy monitoring, database isolation discipline. - Technical coordination with a client's engineering or IT counterpart when integrations come up, mostly async, working alongside my Operations Coordinator to confirm our deployment fits their environment and their data requirements map to our build. I'm on the key calls with you. A word on how we work, because it matters more than the resume: This is a live production platform with real client data, so we run disciplined change-management: one engineer in the repo at a time, backups before any risky operation, feature branches only, and clear rules about database isolation and what may live in shared code. These practices exist for concrete reasons, which I'm glad to walk through on a call. I need someone who treats these guardrails as the job, not as friction. Calm, careful, and reversible beats fast and clever here. You're a good fit if you: - Have 5+ years across full-stack and platform/infrastructure engineering. You can debug a React rendering bug in the morning and reason about a Postgres migration in the afternoon. - Are deeply hands-on with the stack: Next.js 15+, React, TypeScript, Vercel, Railway, Postgres, Prisma, and Clerk (or comparable auth). - Have owned a multi-app monorepo end to end, including the application code, not only the deploy pipeline. - Stay calm under incident pressure and have a track record of fast production fixes that stay fixed. - Have real discipline around data integrity: backups, safe migrations, and knowing when not to run a command. - Are comfortable working in a codebase built with agentic AI tools. Much of this platform is built with Claude Code and similar agents, so you need to work with that grain, not fight it. Hands-on Claude Code or Cursor experience is a strong plus. - Can hold a clear technical conversation with another company's engineer or IT contact, calmly and professionally. You won't own the client relationship, but you will be the technical voice on integration syncs. - Communicate clearly in English. - Are in a US or Latin America timezone with real overlap with US East Coast hours. Engagement terms: - Start: immediately. This is time-sensitive. I have a hard client launch at the end of June and I need someone who can start this week and move quickly. - Ongoing engagement, 3-month initial term with strong potential to extend. - Billed hourly via Upwork, paid weekly. Roughly 30-40 hours per month at steady state (about 8-10 per week). The next three weeks, through the late-June launch, run materially heavier, then settle into the monthly health pass. Rate: this role has two phases. The launch sprint (now through early July) pays $75-85/hour for intensive, higher-stakes work. Steady state after the launch settles to $55-65/hour for the ongoing health pass. Final rate depends on experience. How to apply: Start your proposal with the word "Warren" so I know you read this. Note on process: for some candidates I may ask for a short paid diagnostic task before finalizing, to see how you triage a real problem. It's paid and quick, and it's not a guaranteed step, since for the right candidate I'm ready to move fast. We look forward to connecting with the right candidate soon. Good luck!