Staff Product Manager, LLMX Platform

MaintainX•San Francisco, CA

8h•Hybrid

About The Position

MaintainX is seeking a Staff Product Manager to own LLMX, MaintainX's internal AI Platform. This platform serves as the runtime, orchestration, evaluation, and observability layer for all AI features developed by product teams. This role is highly influential, as decisions made will impact every agent, AI-powered workflow, and intelligent capability shipped by the company. The ideal candidate will possess a blend of technical depth in AI systems and a commercial instinct driven by curiosity. This is a hybrid role based in either San Francisco or Toronto.

Requirements

10+ years in product management, with meaningful experience in AI platform, ML infrastructure, or developer platform roles.
Experience shipping production AI services that other teams built on, and understanding what it takes to earn and keep their trust.
Informed opinions on model selection, RAG vs. fine-tuning trade-offs, agentic architectures, evaluation frameworks, and latency/cost optimization.
Comfortable serving both internal teams building on the platform and enterprise buyers whose outcomes depend on it.
Strong commercial instinct, connecting platform investment to enterprise differentiation, competitive positioning, and deal cycles.
A system for staying current in LLMs and agentic AI, and the discipline to apply it without chasing every new release.
Data-driven approach: define success in metrics, instrument the platform, and use evals to make quality visible at scale.

Nice To Haves

Experience in industrial operations or asset management.

Responsibilities

Own the LLMX roadmap end to end, and drive adoption and quality against metrics like time-to-first-integration, eval coverage, latency, and cost.
Define the platform primitives, including routing agents, skill orchestration, human-in-the-loop workflows, agent memory, conversation compaction, and the app-wide conversational interface, that make MaintainX's next generation of agentic features possible.
Own the platform's evaluation infrastructure, including online eval monitoring, dataset promotion, and agent gating, so quality scales as the number of production agents grows.
Work proactively across MaintainX's entire product organization to identify what upcoming agentic features will need, before those needs become blockers, including demand that's still nascent.
Shape build vs. buy vs. integrate decisions for models, providers, and infrastructure, staying ahead of the state of the art in LLMs and agentic AI.