Open role

Member of Technical Staff (GPU Performance Engineer)

Apply now

Reka AIUS, UK, Singapore, RemotePermanant

About Reka

Reka trains and serves frontier multimodal models — efficiency at the kernel level is a core advantage.

The Role — Member of Technical Staff (GPU Performance Engineer)

Profile and optimize training and inference workloads across GPU fleets
Write and tune custom kernels (CUDA/Triton), attention variants, and memory-efficient ops
Improve throughput, latency, and cost of serving multimodal models

What we look for

Deep experience with CUDA, Triton, NCCL, or equivalent GPU programming
Understanding of transformer inference optimization (KV-cache, quantization, speculative decoding)
Systems mindset: you measure before you optimize

Location: US, UK, Singapore, Remote · Full-time

Application

Apply for Member of Technical Staff (GPU Performance Engineer)

Reka AI

Zero Hiring processes the information you provide (including your resume and interview responses) on behalf of Reka AI to assess your application. Your application may be screened and scored using automated / AI systems. We keep your data as described in our Privacy Policy and you can exercise your privacy rights at privacy@zerohiring.com.

Some screening decisions are made with the help of automated systems. You have the right to request human review of a decision that affects you — contact privacy@zerohiring.com.

Read our Privacy Policy

First nameLast name

EmailPhone number

Resume file

By providing your phone number and checking this box, you agree to receive text messages at that number from Reka AI, including messages related to your application(s), other job opportunities, and work-related messages if you are hired. You can opt out at any time by replying STOP. Message and data rates may apply.*