Software Engineer – AI Inference Engine

FriendliAI•San Francisco, CA

33d

About The Position

We are seeking a highly technical Inference Engine Engineer to optimize the performance and efficiency of our core inference engine. In this role, you will focus on designing, implementing, and optimizing GPU kernels and supporting infrastructure for next-generation generative and agentic AI workloads. Your work will directly power the most latency-critical and compute-intensive systems deployed by our customers. We are looking for an exceptional engineer with a strong foundation in GPU programming and compiler infrastructure. The ideal candidate enjoys pushing performance boundaries and has experience supporting production-scale machine learning applications.

Requirements

5+ years of experience in production or high-impact research environments
Production-level expertise in Python and C++
Bachelor’s or Master's degree in Computer Science, Computer Engineering, Electrical Engineering, or equivalent
Experience developing machine learning frameworks or performance-critical runtime systems
Hands-on experience writing and optimizing GPU kernels
Hands-on experience profiling GPU kernels
Experience working with generative AI models such as transformer and diffusion models

Nice To Haves

Experience developing machine learning compilers or code generation systems
Familiarity with dynamic shape compilation, memory planning, and kernel fusion
Contributions to inference engines, compilers, or high-performance numerical libraries
Understanding of multi-GPU and distributed inference strategies

Responsibilities

Design and optimize custom GPU kernels for AI (e.g., transformer and diffusion) workloads
Contribute to the development of FriendliAI’s kernel compiler, memory planner, runtime, and other core components.
Collaborate with cloud and infrastructure engineers to ensure end-to-end inference performance
Analyze performance bottlenecks across the software and hardware stack, and implement targeted optimizations
Drive support for new model architectures and tensor compute patterns
Maintain production-grade performance infrastructure, including profiling, benchmarking, and validation tools