Principal Software Engineer, GenAI Platform

Roblox•San Mateo, CA

9h•$328,610 - $362,280•Onsite

About The Position

Foundation AI builds the platforms, services, and infrastructure that enable machine learning and generative AI across Roblox. The Principal Engineer sets technical strategy and direction across areas including model training and inference, LLM services, agent infrastructure, data systems, experimentation, evaluation, and developer tooling. This role is accountable for the architecture and long-term health of systems used by many teams and shapes how Roblox builds AI in partnership with engineering, product, research, data, infrastructure, safety, creator, engine, discovery, and economy teams. The position reports to an Engineering Manager within the Foundation AI organization. The role defines multi-year platform architecture and technical vision; leads complex, ambiguous, high-impact platform work through production; establishes self-service standards and guardrails; and makes foundational trade-offs involving quality, latency, reliability, cost, and responsible data use. It also owns operational standards at scale, including SLOs, capacity planning, on-call, and incident response, while identifying opportunities and risks, aligning stakeholders, and multiplying engineering impact through mentorship and technical review.

Requirements

Undergraduate or graduate degree in computer science or a related technical field, or equivalent practical experience.
At least 8 years of experience building and operating large-scale production software, distributed systems, cloud services, or related infrastructure.
Sustained track record designing and owning foundational systems or platforms with organization-wide or company-wide impact.
Deep expertise in one or more of machine learning, generative AI, distributed systems, data infrastructure, or developer platforms, with the breadth to reason across these areas.
Expert proficiency in one or more of Go, Python, or Rust, and sound judgment in choosing tools for durable systems.
History of setting technical direction across multiple teams and driving large cross-organizational efforts to completion.
Exceptional communication and influence skills, with ability to align engineers, product teams, and leadership around technical strategy.

Nice To Haves

Experience building or operating an LLM gateway, inference-serving, or model-routing layer, including multi-provider failover, rate limiting, and token or cost accounting at scale.
Hands-on experience with vector databases, similarity search, or knowledge graphs, and with RAG or agentic systems such as retrieval pipelines, agent memory, context management, or evaluation of non-deterministic systems.
Experience designing multi-tenant platforms, including usage attribution, metering, spend controls, and self-service onboarding.
Background in GPU infrastructure or ML inference efficiency, including batching, quantization, accelerator utilization, or cost optimization.

Responsibilities

Define the multi-year technical vision and architecture for platforms and services powering machine learning, generative AI, and agentic experiences across Roblox.
Lead high-impact, ambiguous platform problems from framing through production, including multi-provider inference resiliency and scaling information retrieval for agentic use, and be accountable for outcomes.
Establish standards, paved paths, and guardrails that make the platform self-service for teams, including multi-tenant usage attribution and metering, built-in spend controls, and ticket-free onboarding.
Make foundational trade-offs in quality, latency, reliability, cost, and responsible data use, treating instrumentation and observability as first principles.
Own platform operational standards at scale, including SLOs, capacity planning, on-call, and incident response; design for graceful degradation and fast recovery.
Identify high-impact opportunities and risks and build alignment across engineering and leadership.
Mentor engineers, contribute to design and code reviews, and set the bar for engineering excellence.