On-Device ML Engineer — Equity & Systems Optimization

Base Compute; Melbourne; On-site; AUD 120,000 - 180,000; Full time

Benefits offered by this job

Equity

Competitive salary

Job summary

Base Compute, an AI inference lab focused on on-device intelligence, is hiring an ML Engineer to own and optimize the entire inference stack. You will drive performance across CPU/GPU, write production-ready systems in C++, and shape hardware- and software-level bets for edge platforms.

Applicants should have 3+ years in ML or systems programming, strong GPU know-how, and a track record of turning research into scalable infrastructure.

Qualifications

• 3+ years of experience in ML engineering or systems programming (Rust, C/C++).
• Expertise in GPU programming and hardware optimization across platforms (CUDA, ROCm, Metal).
• Solid understanding of modern LLM architectures and optimization techniques (quantization, decoding).
• Strong ownership and autonomy: translate research into production-ready infrastructure.
• Good communication: explain complex decisions and document systems clearly.

Responsibilities

• Develop and scale our on-device inference engine and runtime.
• Optimize kernels and leverage hardware features across architectures.
• Build low-latency serving runtimes in C++ for routing, batching, and decoding strategies.
• Profiling to remove bottlenecks from memory to kernel interleaving.

Skills

ML engineering

Systems programming

Rust

C/C++

GPU programming

Hardware optimization

Ownership

Communication

Tools

CUDA

ROCm

Metal

Triton

Torch.compile

Custom operators

Job description

Base Compute, an AI inference lab focused on on-device intelligence, is hiring an ML Engineer to own and optimize the entire inference stack. You will drive performance across CPU/GPU, write production-ready systems in C++, and shape hardware- and software-level bets for edge platforms.

Applicants should have 3+ years in ML or systems programming, strong GPU know-how, and a track record of turning research into scalable infrastructure.