Full Stack Engineer

Most enterprise SaaS models a workflow. This one models a factory — what actually happens inside an industrial production process, unit process by unit process. Messy manufacturer data on one side, reporting a manufacturer can put its name to on the other, and increasingly an agent layer in between doing the extraction and mapping. Our client is an early-stage B2B SaaS company in New York building the data layer for materials manufacturing. Their customers are large industrial manufacturers. The platform ingests raw production data — input materials, production runs, energy, water, waste — and turns it into structured analysis and the verified reporting those manufacturers are increasingly required to produce. Today that work is slow, manual and consultant-driven. They are backed by a substantial seed round, signing multi-year enterprise contracts, and growing a small engineering team that sits directly with the founders. What you'll do: Own features end to end — data models, Python APIs, React UI, and the agentic workflows that sit on top of bothBuild agent pipelines that assist data ingestion, automate client setup, and show customers how an analysis was reachedMake LLM-backed extraction and mapping reliable enough for production — it has to fail safely and be inspectableEvolve backend services, data models and AWS infrastructure ahead of growthWork directly with founders and domain experts on where agents belong in the customer workflow, and where they don't What they're looking for: 6+ years building production software, ideally enterprise-grade B2B SaaSStrong Python on the backend — FastAPI, Pydantic, SQLAlchemy or equivalentReal React and TypeScript on the frontend. Genuinely full-stack, not backend with a little UIAWS in production — ECS, Lambda, SQS, RDS, S3 — plus Docker and CI/CDPostgreSQL and SQL fluency. DynamoDB and OpenSearch are a plusProduction experience with LLM-backed features and agentic workflows is strongly preferredExperience in startup or scale-up environments, where the playbook isn't written yet The team is small, deliberately anti-meeting, and async-first. They want someone who writes clearly, moves decisions forward in writing, and works comfortably when the problem is still loosely defined. Curiosity matters more than pedigree here — the people who do well find modeling an industrial process at the unit level genuinely interesting, and care whether a manufacturer actually uses what they shipped. Strong opinions on code quality are welcome, held loosely enough to change when the problem demands it. Where you went to school is not a factor. Why join? Small team, founder-adjacent, real influence over product and technical directionAgentic AI as a core product surface, not a bolt-on demoA domain with physical stakes and genuinely hard modelling problemsEarly enough that equity is meaningful Benefits: Competitive salary and equity. Company card for office food and snacks. Team offsites. Relocation support if you're not in NYC. Compensation: $170,000–$200,000 base salary, plus equity.