Job Description We’re building intelligent, autonomous AI agents that can reason, plan, and act across real-world workflows. These agents go beyond chat — they take meaningful action, work across tools and systems, and operate in complex, high-stakes environments. As our Product Manager for OEM Engineering, you’ll lead the definition and delivery of early use cases for agentic AI. You’ll turn cutting-edge agent capabilities — like long-term memory, tool use, and planning — into applied, valuable products. You’ll be use-case–driven, but platform-aware — working closely with engineering and research to ensure what we build is feasible, scalable, and aligned with the underlying agent architecture. Use Case Discovery & Validation Identify and prioritize agent use cases with high ROI: operational automation, coordination agents, internal task delegation, etc. Work with users and stakeholders to understand their workflows, challenges, and trust boundaries. Prototype and iterate agent behaviors in production-like environments to validate fit and reliability. Product Design & Execution Define the agent experience — from delegation to monitoring to escalation — for real-world scenarios. Translate use cases into clear product specs: user flows, system interactions, prompts, safety requirements, and interfaces. Own the roadmap and ship iteratively: starting small, learning fast, scaling wisely. Drive OEM partnerships and product strategy to integrate and scale AI agents within enterprise software ecosystems. Cross-Functional Collaboration Collaborate with engineering to align product needs with agent platform capabilities (planning, memory, tool orchestration, retries, etc.). Partner with AI research to bring experimental features into real-world settings. Ensure the platform supports the complexity of applied, multi-step, tool-using agents. Strategy & Metrics Define success metrics around outcomes — time saved, accuracy improved, failures recovered — not just clicks or sessions. Build product foundations that allow for vertical reuse, system generalization, and long-term scalability.