Member of Technical Staff (Software Engineer, Model Platform)

Perplexity is looking for a deeply technical software engineer to own and evolve our model serving platform, which is responsible for connecting our products and research systems with our inference layers. This mission-critical system gives teams one fast, reliable interface across providers. You will work at the intersection of distributed systems, AI products, and infrastructure. You will make it dramatically easier for teams across Perplexity to adopt new models, run rigorous experiments, and ship dependable products. The ideal candidate combines sound technical judgment forged in production systems with active curiosity around the rapidly changing capabilities and failure modes of frontier models. You will design core abstractions, lead high-conviction technical decisions, and write production Rust and Python for a service that serves hundreds of millions of model requests per day.

Responsibilities

Own and evolve our model serving platform; Connect products and research systems with inference layers; Provide teams one fast, reliable interface across providers; Work at the intersection of distributed systems, AI products, and infrastructure; Make it dramatically easier for teams to adopt new models, run rigorous experiments, and ship dependable products; Design core abstractions and lead high-conviction technical decisions; Write production Rust and Python for a service that serves hundreds of millions of model requests per day.

Requirements

5+ years of experience building and shipping large-scale systems and services with high reliability, availability, and performance.; Strong software engineering craft, with special care for developer experience and designing future-proof systems that can evolve along with a rapidly advancing AI landscape.; Proficiency with a general-purpose backend language (Python, Rust, C++, Go) and cloud infrastructure (AWS, Kubernetes).; BS, MS, or PhD in Computer Science, Engineering, or related fields (or equivalent experience).

Nice to have

Sound technical judgment forged in production systems; Active curiosity around the rapidly changing capabilities and failure modes of frontier models.