Senior Machine Learning Operations Engineer, Agentic Systems Platform Engineering

Paramount•New York, NY

About The Position

We are looking for a Senior MLOps Engineer (IC3) to join the Platform Engineering pod. Your mission is to build the Agentic Systems Platform—the runtime environment where autonomous agents live, reason, and collaborate. You will lead the development of the MCP (Model Context Protocol) Gateway, routing logic, and the orchestration support required for multi-agent personalization systems. In this role, you will bridge the gap between traditional ML serving and autonomous agency. You will lead the design of agentic runtimes using GKE and Apigee, utilizing the ADK (Agent Development Kit) and A2A (Agent-to-Agent) protocols to enable complex workflows. Your challenge is to manage the "thinking time" and cognitive overhead of agentic systems, ensuring they remain responsive and cost-effective in a production environment.

Requirements

4–8+ years in MLOps or Platform Engineering
Experience with API Gateways (Apigee) and K8s (GKE)
Background in Agentic Orchestration

Nice To Haves

Familiarity with A2A protocols
Experience managing long-running stateful workflows

Responsibilities

Lead Platform Design: Architect the production runtime for agentic systems, focusing on low-latency tool-use and state management.
Own the MCP Gateway: Design and implement the routing and security layer for agentic tool-use via Apigee.
Standardize Agentic Workflows: Develop orchestration patterns using ADK and A2A that are adopted by pods across the AMLG.
Performance Management: Monitor and optimize agentic "thinking" cycles to balance reasoning depth with production latency constraints.
Mitigate System Risks: Proactively identify and fix failure modes unique to agentic systems (e.g., tool-use loops or context window overflows).