Senior Principal Product Manager
Zendesk
- Location
- Melbourne, Australia
- Work model
- On-Site
- Level
- Principal
- Posted
- Jul 15, 2026
Skills
About this role
Job Description
Senior Principal Product Manager, AI Platform, Australia (Melbourne) This role is based in Melbourne and requires you to work from office at least 2 days a week.
Job Description
Zendesk is seeking an experienced Principal Product Manager to lead product for our internal AI Platform — the foundational platform that powers every AI feature across Zendesk. This platform spans the full agent lifecycle (build, test, publish, operate, improve) and provides the shared capabilities, infrastructure, SDK, and developer tooling that all vertical AI teams build upon. The right candidate will demonstrate strong technical, product and strategic thinking with a customer-centric mindset, excellent prioritization skills, and flawless execution ability. You will collaborate cross-functionally to drive product vision and strategy, while ensuring successful execution in this technical domain.
What You'll Do
Own the end-to-end product strategy and roadmap for Zendesk's internal AI Platform, a layered stack spanning foundational infrastructure through to developer-facing tooling: Agent Development Kit (ADK) & SDK — the primary interfaces for vertical teams to build, compose, and deploy AI agents, including runtime protocols, registry clients, CLI tooling, and YAML-based agent authoring. Platform Capabilities — durable execution, agent memory, CI/CD for agents, A/B testing, agent registry (skills, tools, agents, guardrails), and governance across the full agent lifecycle. Evaluation & Observability — a layered observability stack from raw agent traces through to customer-facing evals, conversation simulation, QA integration, and A/B testing frameworks. Infrastructure — orchestration (Temporal), eval tooling (BrainTrust), feature management (GrowthBook), agent sandboxing (MicroVMs, FireCracker, WASM), and foundational compute, storage, and event systems. Drive platform adoption across vertical AI teams — meeting teams where they are and enabling progressive adoption of shared capabilities to help them meet their goals Lead the development of Zendesk's A2A (Agent-to-Agent) layer and agent registry, enabling connection, reuse, and governance across existing agents and applications. Conduct internal user research and external industry research to define a forward-thinking roadmap that sets Zendesk up to accelerate AI development in the short and long-term. Collaborate with stakeholders to translate evolving machine learning and AI platform needs into clear, actionable requirements that balance innovation, reliability, and cost efficiency. Manage cross-team collaboration, prioritization, and trade-offs in a complex, rapidly evolving AI product environment.
What You'll Bring
5+ years of product development experience, with significant experience in developer platforms, AI/ML infrastructure, or internal tooling at scale. Deep understanding of AI/ML concepts, model lifecycle, agent architectures, distributed execution, and platform engineering. You have a technical background and 10 yrs product management experience. Proven ability to drive platform adoption and developer experience for internal customers — you understand what makes engineers want to use a platform vs. build their own. Experience leading multi-layered platform initiatives with strong strategic vision, from low-level infrastructure through to developer-facing SDK/API design. Demonstrated success managing experimentation and evaluation platforms, developer-facing APIs, agent frameworks, or AI developer ecosystems. Exceptional communication and stakeholder management skills enabling influence across engineering, data science, security, and product teams. Skilled in balancing competing priorities, managing ambiguity, and driving consensus in complex environments. Passion for building scalable, reliable, and easy-to-use AI infrastructure products that empower internal and external users. Nice to have skills Experience with building agentic features and