Senior Product Manager, ML Modeling & Platform

About the role

This role manages the dual responsibility of Klaviyo's predictive modeling roadmap and the underlying infrastructure that powers it. You will work directly with the engineering team in Palo Alto to bridge the gap between technical platform capabilities and tangible customer outcomes.

Key facts

What you'll do

• Define the product strategy for core predictive and generative models, including churn prediction, product recommendations, audience optimization, and smart send time.
• Manage the roadmap for the internal ML infrastructure, covering DART (Ray-based job platform), MLflow, model serving, feature pipelines, and Prefect workflows.
• Establish and maintain monitoring and evaluation frameworks to ensure model quality and prevent regressions.
• Collaborate with engineering leadership to determine build versus buy strategies for the ML stack.
• Align platform and modeling investments with business and customer goals by coordinating with go-to-market and success teams.
• Optimize ML workloads for reliability, cost efficiency, and observability across 200,000+ customers and billions of events.
• Improve developer velocity by reducing friction in the model shipping and production maintenance lifecycle.
• Guide the evolution of the ML platform to support the specific demands of agentic products and LLM tooling.

Requirements

• 5+ years of experience in product management, specifically with ML, AI, or data platform products in production environments.
• Working knowledge of ML systems, including feature stores, training pipelines, experiment tracking, and model serving.
• Ability to translate technical metrics, such as training throughput, into business narratives for leadership and stakeholders.
• Experience operating as the sole product manager within a highly technical engineering team.
• Capability to manage tradeoffs between inference latency, training costs, model quality, and developer speed.
• Ability to prioritize long-term infrastructure investments alongside immediate product requirements.

Skills & tools

• Ray (DART)
• MLflow
• Prefect
• Distributed training
• Inference infrastructure
• Feature pipelines

Practical notes

Klaviyo encourages candidates to apply even if they do not meet every requirement listed. The company values diverse backgrounds and perspectives.