ML Software Engineer; L6 — Platform Systems, AIMS Engineering

Role Overview

AI for Member Systems (AIMS) powers the recommendation, search, and personalized experience for over 300M members. The stack is mature but must evolve to support new model paradigms, tighter cost and efficiency expectations, and operational maturity s Staff ML Software Engineer will own the end‑to‑end modernization of the AIMS AI/ML stack, build observability and cost infrastructure, and define its long‑term architectural evolution.

• Define the end‑state architecture for the modernized AIMS AI/ML stack: organization, contracts, and migration path across training pipelines, AI frameworks, and data infrastructure.
• Drive end‑to‑end migration of AIMS AI/ML systems onto a modern, Python‑native platform, coordinating across multiple AIMS teams and external platform partners, with dozens of production models in flight.
• Build migration tooling and shared abstractions that reduce the cost of adoption for individual teams, so modernization does not require each team to solve the same problems independently.
• Own scalability across training throughput and data pipelines, ensuring AIMS AI/ML systems stay performant as model complexity and member traffic grow.
• Design and build observability systems that give AIMS ML practitioners deep visibility into model behavior, training pipeline health, serving latency, and data quality, making issues detectable and diagnosable before they become incidents.
• Identify and drive cost optimization across AIMS training and serving infrastructure, developing frameworks and tooling that make compute efficiency a first‑class concern, not an afterthought.
• Architect reliability improvements across the AIMS AI/ML stack, reducing toil, improving on‑call ergonomics, and setting the standard for operational excellence across the org.
• Prototype and product ionize GenAI‑powered tooling for anomaly detection, root cause analysis, and operational automation, applying LLM‑based systems to the problems of AI/ML reliability and cost at scale.
• Surface systemic cost, reliability, and migration gaps by embedding with AI/ML teams across AIMS, and translating their friction into concrete engineering investments with org‑wide leverage.
• Set technical standards for the modernized stack and raise the engineering bar across AIMS through design reviews, architectural guidance, and leading by example.
• Own the long‑term architectural evolution of the AIMS AI/ML stack — continuously evaluating emerging infrastructure patterns, model paradigms, and platform capabilities, and translating them into a forward‑looking roadmap before they become urgent migrations.

• Significant experience designing, building, and operating large‑scale production AI/ML systems, including training pipelines and familiarity with model serving and online inference at high‑traffic scale.
• Hands‑on experience migrating production AI/ML systems across technology generations; you have done this before and understand where it goes wrong.
• Strong software engineering fundamentals with deep Python expertise and working proficiency in at least one JVM language (Scala or Java).
• Proven track record of improving AI/ML system reliability, reducing infrastructure costs, and improving operational scalability.
• Experience building observability and monitoring systems for AI/ML workloads; you understand what good visibility looks like across training, serving, and data pipelines.
• Strong distributed systems background, including large‑scale batch processing and real‑time serving infrastructure.
• Collaborate with partner teams to drive cross‑functional technical programs, setting direction, managing dependencies, and building consensus without formal authority.
• High technical judgment: able to identify common patterns, build reusable frameworks, and make pragmatic calls on what to migrate, what to rewrite, and what to leave alone.
• Comfortable operating without full information; you can scope a problem, define an approach, and course‑correct as you learn more.

• Experience with compute and cost optimization for AI/ML workloads at scale, including capacity management and efficiency tooling.
• Hands‑on…