Staff Full Stack Engineer

Knit is an AI-native consumer research platform automating primary research for enterprise brands. The role helps scale the NYC engineering team from current size to 20+ within 18 months. As a Staff Full Stack Engineer, you'll architect and own the core agent orchestration engine, building multi-agent LLM workflows with tool use, memory management, and AI-native UX design. Day-to-day you'll develop end-to-end features spanning design through production deployment, design and maintain multi-agent workflows, create React/Next.js interfaces enabling human-AI collaboration, build internal experimentation infrastructure for prompt and agent testing, own observability, testing, and deployment safety, and mentor team members. Stack: Python, TypeScript, React, Next.js, FastAPI, MongoDB, S3, Redis, Pinecone, AWS, Docker, Kubernetes, GitHub Actions. Required: LLM and agent systems experience, evals and prompt tooling, RAG and embeddings work, vector databases, event-driven systems and workflow engines, observability (OpenTelemetry, Prometheus, Datadog), enterprise security (SOC 2, GDPR). Hybrid — 3+ days/week in Midtown Manhattan.