Principal Software Engineer United States Washington Redmond

Job description

Design and ship AI systems end‑to‑end—from exploration to production—spanning LLM‑based services, retrieval/grounding pipelines, and robust evaluation/observability. Build agentic and tool‑using workflows that reason across data and services, optimizing for safety, latency, and cost at cloud scale (Azure preferred). Collaborate across research and product teams to adapt/fine‑tune models (LLMs/SLMs, multimodal) for real customer scenarios. Lead technical direction across features and services; mentor engineers; champion high engineering standards, reliability, and accessible, inclusive experiences. Uphold Responsible AI, secure and compliance practices in design, data handling, and experimentation. Participate in on‑call/operational excellence as part of a healthy, sustainable engineering culture; lead by example. 8+ years building and shipping production software/services including 3+ years with modern ML (e.g., transformer‑based LLMs) and applied AI/ML systems. Coding skills in one or more: Python, C#, or C++—with design/architecture experience for cloud services (Azure preferred). Hands‑on experience taking LLM/ML features from prototype to production (evaluation, safety, telemetry, and iterative improvement). Ability to influence across teams and drive ambiguous, multi‑partner initiatives to outcomes. Bachelor's degree in computer science/engineering or equivalent practical experience. Experience with agentic systems, model/tool orchestration, and/or retrieval‑augmented generation (RAG). Expertise with Azure AI, vector stores, prompt/response evaluation, and performance optimization for LLM workloads. Track record of shipping applied research to production and partnering with research teams on custom/fine‑tuned models. Familiarity with Responsible AI practices and inclusive design/accessibility standards.