Product Manager, Inference

LiveKit

4d•$225,000 - $265,000

About The Position

LiveKit is seeking a Product Manager to own LiveKit Inference, a product that provides developers with managed access to various models for voice AI through a single integration. This is a key revenue-generating product within the company, and the Product Manager will be responsible for establishing its foundation and future direction. The role involves setting the vision and roadmap, managing commercial relationships with model providers, overseeing model hosting and optimization, and ensuring LiveKit stays competitive in the rapidly evolving AI model landscape. The goal is to make LiveKit Inference the preferred choice for voice AI workloads, offering advantages in ease of use, speed, cost, and quality compared to direct integrations with model providers.

Requirements

8+ years in product management, at a senior or staff level.
Experience owning a developer-facing API, platform, or AI/ML product.
Technical depth to reason about model serving, routing, latency, and quality tradeoffs, and to gain the trust of engineers.
Experience owning a product where cost, performance, and vendor/provider tradeoffs were central, and the ability to keep a product current as the landscape shifts.
Ability to influence cross-functional teams through clarity, trust, and good judgment.
Strong communication skills, including clear writing, effective presentation, and the ability to navigate strategic and technical discussions.
Comfort with ambiguity, prioritizing effectively, and moving quickly with incomplete information.

Nice To Haves

Experience building a model gateway, inference API, model-provider product, or AI/ML platform.
Familiarity with model benchmarking and measurement of performance and quality (e.g., TTFT).
Familiarity with the STT, TTS, LLM, and real-time model space and top providers.
Background at a developer tools, API infrastructure, or AI/ML company.
Familiarity with voice AI, real-time communication, or telephony.
Experience with usage-based pricing models and cost or margin optimization.

Responsibilities

Own the LiveKit Inference roadmap, setting a clear vision and a sequenced plan for product development in partnership with engineering.
Ship a model layer that becomes the default choice for developers, making model access through LiveKit easier and better than direct provider integration.
Stay current with the model landscape, deciding which models, providers, and modalities to add to keep LiveKit competitive.
Grow the value delivered and captured by consolidating model spend onto LiveKit, enhancing customer value without increasing costs.
Engage deeply with customers and data to understand production model usage and inform product priorities.