Principal Software Engineer, Agent Infrastructure
Roblox · San Mateo, CA, United States
About this role
**Principal Software Engineer, Agent Infrastructure** Roblox is building the tools and platform that empower our community to bring any experience they can imagine to life—reimagining how people come together from anywhere in the world, on any device. ## Why Roblox & Agent Infra The Agent Infra team builds the infrastructure that runs autonomous AI agents at Roblox for millions of creators and for thousands of engineers shipping production code daily. Agent workloads break traditional infrastructure assumptions: they run for hours (not milliseconds), are non-deterministic, execute code nobody wrote, and require real credentials against real systems. Your work will make this safe, durable, and cost-effective at Roblox scale. ## You Are - **Demonstrated experience:** 8+ years building and operating backend, distributed, or infrastructure systems in production (including systems you designed that held up at scale for years) - **Deep expertise** in at least one: workflow orchestration, scalable backend infrastructure, agentic runtimes, or secure sandboxing & access - **Hands-on:** you spend most of your time building and write production code in **Go, Python, Rust,** or similar - **Agent-native:** you’ve experimented with swarms/loops/orchestration, have strong opinions on harness & model choice, and continuously evolve your agentic workflow - **Experienced with agents in production:** you’ve run agents against real systems, observed failure modes, and know what the underlying infrastructure should do - Bachelor’s degree or equivalent experience in CS/CE or a similar technical field ## You Will - **Set the technical direction:** Own the design and implementation of Roblox’s core agent infrastructure and scale it for internal and external customers - **Drive reliability and security:** Keep long-running agents alive through deploys, restarts, and outages; run untrusted code in sandboxes that protect credentials and control network access; define **SLOs** and observability - **Raise the bar:** Keep learning and experimenting with agentic workflows and bring what works to the team and the wider Foundation AI org - **Align across the org:** Partner with Security, Engineering Efficiency, and Infrastructure to ensure the agent platform and dependent systems move in the same direction - **Multiply the team:** Mentor senior engineers, lead design reviews, and push how far agents can build the agent platform itself ## Nice to Have - Security experience with systems that run untrusted code (sandbox/container isolation such as **gVisor, Firecracker, microVMs**), Kubernetes hardening, network segmentation, and egress control - Experience with **Temporal**, **AWS**, and self-hosted models - Experience building/deploying production systems on **Kubernetes** using **GitOps** and **Argo CD** - Open source contributions or experience building your own product with real user traction ## Compensation & Location Notes - **Annual Salary Range (HQ in San Mateo, CA):** **$278,530 — $345,040 USD** (actual base pay depends on factors like background, experience, location, and market demand) - Full-time employees are also eligible for **equity compensation** and **benefits** - Office-based roles are **onsite Tuesday/Wednesday/Thursday**, with optional presence on **Monday/Friday** *Roblox is an equal opportunity employer and provides reasonable accommodations during the recruiting process.*
Listing freshness
CronJobs last confirmed this listing 1h ago. If its source stops confirming the opening for seven days, this page is removed from active inventory.