AI Orchestrator

Role overview Design and deliver the systems that coordinate AI agents across real business workflows. You’ll decide how agents divide tasks, share context, use tools, and involve people, while balancing output quality, cost, and response time. The work spans architecture, implementation, evaluation, and direct collaboration with clients.

Responsibilities - Design multi-agent workflows, including task sequencing, branching, fallback paths, and recovery behavior. - Select models for individual tasks and manage trade-offs among quality, cost, and latency. - Build end-to-end evaluation methods to assess the quality of complete workflows. - Add safeguards such as content filters, confidence thresholds, hallucination checks, and human review steps. - Connect language models to APIs, databases, tools, and enterprise systems. - Explain design choices to clients and adapt systems as their needs and conditions change.

Requirements - 3–7 years in software engineering, including 1–2 years building production systems with language models or multi-agent approaches. - Familiarity with orchestration frameworks and patterns, such as LangChain, LangGraph, CrewAI, or tool-use APIs, and judgment about when to apply them. - Practical understanding of reliable workflow design, retries, context windows, token budgets, and caching. - Experience with retrieval-augmented generation, prompt chaining, and context management across multiple steps. - Product judgment focused on whether systems work for their intended users, as well as on evaluation results. - Clear communication skills, including the ability to explain system designs to non-engineers.

Benefits and work setup - Remote-first role open in the US and Canada, with a distributed team and a Seattle home base. - Competitive salary, equity, and benefits.