Staff Engineer, Agent OS & AI Orchestration

Boson AI; Palo Alto (CA); On-site; USD 150,000 - 400,000; Full time

Benefits offered by this job

401k plan

HSA

FSA

Free meals

Job summary

Boson AI in Palo Alto is seeking an experienced engineer to evolve the core Agent OS—a high-performance, resilient platform powering Boson agents. This role spans dialog & policy engine, memory, tool calls, ReAct-style reasoning, and complex multi-step workflows.

You will own the dialog engine, memory systems, and orchestration frameworks, building a scalable internal SDK/Framework and integrating with RAG, tool calls, and secure runtimes.

Qualifications

• 3+ years of backend engineering and distributed systems experience.
• Experience architecting/operating production-grade agentic systems.
• Familiarity with orchestration patterns like RAG, ReAct, multi-step planning.
• Security-focused development including authn/authz and data protection.

Responsibilities

• Own the core dialog & policy engine and define state machine for agent states.
• Design and maintain high-performance context and memory systems with low latency.
• Architect robust agentic orchestration patterns for multi-step workflows.
• Develop internal frameworks and DSLs similar to LangChain/LlamaIndex with safe APIs.
• Oversee voice/runtime components for low-latency, reliable operation.
• Ensure robust tooling, authentication, rate limiting, and safe I/O between agents and services.
• Shape platform APIs and abstractions for scalable usage by teams.
• Explore and prototype on-device personalization and privacy-preserving approaches.

Skills

Backend engineering

Distributed systems

Agentic systems

Technical leadership

Tools

LangChain

LlamaIndex

Job description

Boson AI in Palo Alto is seeking an experienced engineer to evolve the core Agent OS—a high-performance, resilient platform powering Boson agents. This role spans dialog & policy engine, memory, tool calls, ReAct-style reasoning, and complex multi-step workflows.

You will own the dialog engine, memory systems, and orchestration frameworks, building a scalable internal SDK/Framework and integrating with RAG, tool calls, and secure runtimes.