Pearl Research is an AI lab building custom algorithms and systems that make frontier LLM inference dramatically faster and cheaper. Alongside our research, we run a live serverless platform serving open-weight models at scale across heterogeneous hardware.
We’re hiring an Inference Platform Engineer to build and operate our serving stack end-to-end (request routing, distributed caching, gateway services, and the model runtime) and debug tough live issues from cross-region latency down to subtle model behavior. We’re looking for strong software engineering skills and an eagerness to learn, not prior ML experience or familiarity with a specific stack.
You’ll join a small, stacked Tel Aviv team of IOI/IMO medalists, CS professors, and engineers from NVIDIA, Google, and Amazon, backed by Lightspeed, Ribbit, and leaders from OpenAI and DeepMind. Customer traffic is ramping fast, and you’ll work right next to researchers shipping straight to production with high ownership and zero bureaucracy.
Send your CV to careers@pearlresearch.ai
