About the role
You will own the execution layer for production-ready agents, including orchestration, parallel execution, delegation, sandbox execution, context engineering, and failure recovery. You will define product roadmaps, partner with modeling and engineering on architecture and evaluations, gather customer failure cases, and guide adoption of agent ecosystem standards.
Responsibilities
- Define and own the roadmap for the agent harness
- Own agent-loop, context-engineering, tool-orchestration, sandbox-execution, and sub-agent-delegation capabilities
- Serve as the interface between engineering and modeling teams
- Own the agentic evaluation framework
- Gather enterprise customer feedback on agent failures and translate it into requirements
- Define agentic-modeling goals, priorities, capabilities, and evaluations
- Track the agent ecosystem and guide architecture adoption decisions
Requirements
- 5+ years of product management experience in research, agentic AI systems, or applied ML products
- Understanding of agentic harnesses
- Understanding of LLM agent architectures, multi-agent systems, tool-augmented reasoning, memory, retrieval, orchestration, RAG, and long-horizon execution
- Understanding of agentic evaluation design
- Ability to contribute to implementation-level architecture decisions
- Ability to work across ML research and engineering architecture
- Experience shipping platform-layer products
Benefits
- Weekly lunch stipend
- Health and dental benefits
- Mental-health budget
- RRSP matching, 401K, and pension scheme
- Parental-leave top-up for up to six months
- Arts, culture, fitness, wellness, quality-time, and workspace-improvement benefits
- Education and learning stipend
- Six weeks of paid vacation
- Travel budget for other offices and annual company offsite
- Coworking benefit
- Home-office stipend
Requirements
- 5+ years of product management experience in research, agentic AI systems, or applied ML products
- Understanding of agentic harnesses
- Understanding of LLM agent architectures, multi-agent systems, tool-augmented reasoning, memory, retrieval, orchestration, RAG, and long-horizon execution
- Understanding of agentic evaluation design