Senior Agent Engineer — Open-Source AI Orchestration & Equity

Doist; San Francisco (CA); On-site; USD 200,000 - 300,000; Full time

Benefits offered by this job

Equity

Health insurance

401(k)

Job summary

Doist seeks a Senior Agent Software Engineer to help build the agent systems behind AI workflows, including runtime, orchestration, tools, evals, and product surfaces. You’ll enable agents to understand context, use tools safely, recover from failures, and complete complex tasks across real environments.

You will prototype agent use cases and turn the best into production-quality systems, with a strong emphasis on model-agnostic routing and cost/performance trade-offs.

Qualifications

• 5+ years software engineering with leadership or staff-level experience.
• Strong background in frontend/backend features across systems.
• Experience with TS/Node.js, Python, Go, or Rust in production.
• Familiar with LLM patterns: tool calling and context management.
• Able to prototype and productionize agent use cases.

Responsibilities

• Build core systems for agentic workflows—planning, tool use, context management, and memory.
• Design infrastructure for agents to invoke tools, edit files, run commands, and call APIs.
• Improve agent quality via evals, traces, and production debugging.
• Develop model-agnostic routing, caching, streaming, and cost/perf strategies.
• Collaborate with product, design, research, and open-source contributors to ship features.
• Own major technical initiatives from architecture to launch and monitoring.

Skills

Leadership experience

Product engineering

Systems thinking

Break down complex goals

Decision making

Tools

TypeScript/Node.js

Python

Go

Rust

LLM tool-calling

Model routing

Agent systems

Job description

Doist seeks a Senior Agent Software Engineer to help build the agent systems behind AI workflows, including runtime, orchestration, tools, evals, and product surfaces. You’ll enable agents to understand context, use tools safely, recover from failures, and complete complex tasks across real environments.

You will prototype agent use cases and turn the best into production-quality systems, with a strong emphasis on model-agnostic routing and cost/performance trade-offs.