Software Engineer (Model Inference)

About the role

We are looking for a Software Engineer to own and evolve our model inference stack, delivering low‑latency, high‑throughput AI services to millions of users. The role combines deep GPU‑accelerated systems work with product‑focused delivery in a fast‑moving AI entertainment company.

Key responsibilities

• Design, build, and operate a high‑throughput inference server that serves tens of millions of generations per day with sub‑200 ms latency.
• Optimize GPU utilization through batching, quantization, and custom CUDA kernels to squeeze maximum performance from our fleet.
• Productionize LoRA and other in‑house models, scaling them to serve tens of millions of users.

Required profile

• 5+ years of experience building large‑scale software, preferably in ML inference or GPU‑accelerated systems.
• Proven ability to take ideas from user research through implementation, iteration, and delivery.
• Strong drive to solve difficult problems and deliver impact at scale.

Required skills

• GPU inference (batching, quantization)
• Serving frameworks such as vLLM, TensorRT, Triton
• Experience with custom CUDA kernels

What we offer

• Top‑of‑market compensation with equity and superannuation.
• Real ownership from day one and fast‑growing scope.
• Company card for food, coffee, tools and an unlimited workspace budget.
• Visa sponsorship and relocation assistance to Sydney.