Node.js and AI Developer

About the role We're an early-stage AI education startup building a voice + text tutoring product for students and are looking for a second developer. The product is live and in active use, and we're looking for a senior full-stack engineer to take ownership of the codebase end to end. This is a long-term, hourly engagement (~30–40 hrs/week to start) with room to grow into a core technical role as we scale. You'll work directly with the founder. You should be comfortable owning features from spec to production deploy, making sound architecture calls on your own, and keeping things shipping without heavy hand-holding. What you'll own A production Next.js 14 (App Router) + TypeScript web application A Python real-time voice agent (WebRTC / LiveKit-style streaming, STT → LLM → TTS pipeline) LLM integration work — prompt engineering, tool-calling, model routing, latency and cost optimization across providers Postgres (hosted) data layer, schema design, migrations Self-hosted deployment — Docker, container-based CI/CD on a Linux VPS (not Vercel/Heroku; you should be comfortable with self-managed infra) A growing content/data pipeline (structured curriculum data, ingestion, validation scripts) Required skills 4+ years professional full-stack experience Strong TypeScript + React/Next.js (App Router) — not just pages-router familiarity Strong Python for backend services and real-time/async work Solid Postgres — you can model data, write non-trivial queries, and manage migrations Docker and comfort deploying/debugging on a self-managed Linux server Experience integrating LLM APIs (OpenAI / Anthropic / open models) in production — prompts, streaming, tool use Git discipline: clean branches, PRs, no force-pushing over teammates Nice to have Real-time audio / voice agents (LiveKit, WebRTC, streaming STT/TTS) Latency- and cost-sensitive LLM systems (caching, model routing, hybrid small/large model setups) Tailwind CSS Experience working solo or as the lead dev on a small team Familiarity with the Indian EdTech / K-12 space (a plus, not required) How we work Async-first, with a few weekly syncs. You'll get clear specs and a well-documented codebase. We move fast and ship often. We value engineers who flag problems early and push back when something's a bad idea. We expect ownership: when you pick up a task, you see it through to a working production deploy and confirm it actually works. To apply Please include: A short note on a real-time or LLM-powered system you've built and shipped (your specific role, the stack, what was hard). Links to relevant work — GitHub, live projects, or a portfolio. Your hourly rate and weekly availability (and your time zone / overlap with IST). To confirm you read this: start your reply with the word "TUTOR". We're reviewing applications on a rolling basis and want to onboard quickly. Looking forward to hearing from you.