Senior Engineer - ML Systems (AI Products)

Aimlroles; Sydney; Hybrid; AUD 180,000 - 240,000; Full time

Benefits offered by this job

Hybrid work model

Job summary

Xero is seeking a Machine Learning Engineer to lead the design and implementation of large-scale, production-grade distributed systems powering AI features for millions of daily users. You will own architecture decisions, reduce technical debt, and drive strategy across the AI Products estate.

You will mentor engineers, uplift data usability, and collaborate with Applied Scientists to productionize ML models and LLM-based features.

Qualifications

• 5+ years building and operating production Python services at scale.
• Own services or pipelines in production, including on-call responsibility and incident response.
• Deep understanding of distributed processing principles (Spark, Dask) and strong SQL capabilities.
• Experience integrating ML models or LLM-based features into production systems.
• Familiarity with ML tooling such as MLFlow, TensorFlow, or PyTorch and data orchestration tools like Airflow or Prefect.

Responsibilities

• Design and implement large-scale distributed production systems for AI features.
• Own architecture decisions, manage technical debt across AI Products estate.
• Mentor junior engineers and collaborate to lift the team's capabilities.
• Collaborate across the organization to enhance data usability and reliability.
• Deploy and operate in production environments on AWS and Kubernetes, integrating ML/LLM features.

Skills

Python at scale

Distributed systems

SQL proficiency

On-call / incident response

ML model integration

LLM tooling

Spark/Dask

Tools

Spark

Dask

Airflow

Prefect

MLFlow

TensorFlow

PyTorch

Job description

The role

You will lead the design and implementation of large-scale, production-grade distributed systems that power AI features for millions of daily users. You'll own the architecture decisions that keep our systems flexible, cost-effective, and robust, direct strategy for distributed systems, and manage technical debt across the AI Products estate.

Beyond architecture, your impact lies in lifting the technical capability of the entire AI Products team. You will champion engineering excellence, mentor junior engineers, and collaborate across Xero to enhance data usability - applying modern AI research, including Large Language Models, only once it can be engineered into reliable, production-grade systems.

The team

You will join the AI Products group, a diverse team of scientists, engineers, product managers, and analysts within our broader Data & Science division. As a Machine Learning Engineer, you'll partner closely with Applied Scientists to build the interfaces and harnesses that safely and reliably transition models from research into production. Together, this collaborative team reduces toil and delivers beautiful, data-driven insights for small businesses.

The team is currently working on

• Designing and building highly scalable, distributed production infrastructure to support generative AI features
• Harnessing tools like Python, SQL, and distributed processing engines such as Spark or Dask to handle web-scale data workload.
• Deploying to production environments running on AWS and Kubernetes Integrating modern Large Language Model technologies into product features once they're production-ready

Where and how you can work

Xero offers a flexible hybrid working model designed to blend the collaboration of office life with the autonomy of remote work. You will have access to our modern office spaces, with expectations around office days and collaborative 'boost days' aligned to help your team connect and ship great code effectively.

Here are some of the things we are looking for

• 5+ years building and operating production Python (or equivalent language) services at scale - this is a software engineering role first; strong system design and coding proficiency are non-negotiable
• A track record of owning services or pipelines in production, including operational/on-call responsibility, incident response, and managing technical debt over time
• Deep understanding of distributed processing principles (Spark, Dask, or similar) alongside strong SQL capabilities
• Demonstrated experience integrating ML models or LLM-based features into production systems - you don't need a research background, but you should be comfortable working alongside Applied Scientists to productionize their work
• Familiarity with ML tooling such as MLFlow, TensorFlow, or PyTorch, and data orchestration tools like Airflow or Prefect, is valued but production engineering depth matters more than research exposure
• Nice to have: prior experience applying or fine-tuning LLMs in a product context, though this is not a substitute for the core software engineering bar above