Rust/Linux Tools Developer for AI Inference DevEx

Fractile; Greater London, West of England; On-site; GBP 90,000 - 130,000; Full time

Benefits offered by this job

Competitive salary

Equity & Ownership

Private Medical, Dental and Vision

Contributory Pension

25 Days holiday plus bank holidays

Life/Critical Illness Insurance

Job summary

Fractile is hiring for a Developer Experience professional in the Software org, focusing on the tooling and workflows that enable efficient inference at scale. You will work on ML compilers, device drivers, runtimes and tooling to empower LLm engineers and optimize performance across the stack.

You will identify opportunities, design tools, and deliver them with minimal disruption to existing workflows while collaborating with multiple teams across the company.

Qualifications

• Experience with inference frameworks such as JAX and/or PyTorch.
• Software development in Python, Rust, C/C++, or a similar language.
• Monitoring and optimising code written in CUDA, ROCm, OpenCL, or similar.
• Monitoring and optimising LLM deployments with 50+ billion parameters.
• Experience working with stakeholders across multiple teams.

Responsibilities

• Identify tooling opportunities and design those tools for the LLM engineer workflow.
• Deliver tooling in a way that fits existing workflows with minimal disruption and effort.
• Understand current state of development for inference hardware to identify opportunities for performance improvements.
• Collaborate across teams to ensure tooling meets cross-functional needs.

Skills

JAX

PyTorch

Python

Rust

C/C++

CUDA

ROCm

OpenCL

LLM deployments

Stakeholder management

Job description

Fractile is hiring for a Developer Experience professional in the Software org, focusing on the tooling and workflows that enable efficient inference at scale. You will work on ML compilers, device drivers, runtimes and tooling to empower LLm engineers and optimize performance across the stack.

You will identify opportunities, design tools, and deliver them with minimal disruption to existing workflows while collaborating with multiple teams across the company.