Software Engineer, Infrastructure

Stack

Responsibilities

• ▹Own CVI's inference deployments
• ▹Operate the GPU infrastructure serving live conversations across multiple providers and regions
• ▹Tune newest GPU generations and cut cold-start and model load times
• ▹Expand the GPU footprint: bring on new providers and regions, stand up clusters on EKS
• ▹Push the uptime bar higher, including SOC2 and security work
• ▹Fix problems wherever they live, including backend or product code
• ▹Rework the deploy pipeline so shipping is fast and predictable

Requirements

• ▹Hands-on GPU inference experience, building reliable systems on GPU cloud providers
• ▹Kubernetes and EKS depth, including routing and scheduling
• ▹Deep AWS experience, turning services into simple, repeatable processes
• ▹A senior track record of ownership: setting technical direction, driving ambiguous work to completion
• ▹Ability to explain complex ideas clearly to engineers and non-engineers alike

Nice to have

• ▹Experience with GCP
• ▹Experience with video streaming infrastructure
• ▹Experience with training infrastructure or LLM serving
• ▹Experience with SOC2 or security compliance

Soft skills

About the company

Tavus is a research lab pioneering human computing: real-time human simulation models that let machines see, hear, respond, and look real. The company is Series B, backed by investors including Sequoia Capital, Y Combinator, and Scale Venture Partners.