Senior Machine Learning Engineer
在澳大利亚，我们可以在任何地方进行远程工作或在阿德莱德或珀斯的混合工业。关键矿物质是清洁能源转型的瓶，找到它们更快就像地质物理一样是一个机器学习问题。在舰队空间，我们的卫星启用勘探技术帮助生成数据和模型以更快，更高效地寻找沉积物的数据。为了继续建造和扩展产品，我们正在寻求一名高级机器学习工程师来帮助建立我们的 ML 运营，开发，成熟和部署这些工作背后的模型和系统。你将与我们的另一个高级 ML 工程师和数据科学团队合作，从笔记本和原型到生产级产品的研究模型，你将拥有这些系统运行的管道，GPU 环境和基础设施。这是一个重点在扩大我们的 ML ops 与我们的高级 ML 技术人员一起的工作角色。你将设计无泄漏安全评估，验证不确定性校准是新的数据登陆时有效的，并让每个模型的谱系可复制自配置到预测。你还将拥有我们的 ML 实验平台：基于 ML 流跟踪和模型登记，运行身份和来源公约以及培训工作的环境和依赖性。平台团队共享基础设施模块时，你将重新使用它们，并贡献新的模式。你将会做什么 · 与数据科学团队合作，将研究模型转化为生产准备的产品，以代码质量，可扩展性和最佳实践而不是后续装修。·构建和维护强大的，高效的 ML 和数据管道及其背后的基础设施，包括配置 GPU 环境用于可扩展的 ML 系统的重训周期：·自有模型质量：· 拥有 ML 实验平台的端到端：ML 流量跟踪和模型登记，运行身份与来源公约以及环境和依赖培训工作。· 与平台团队合作，在现有的共享基础设施模块重新使用并贡献新的模式。· 用单元测试和自动化端到端测试支持的清洁可检测 Python （SOLID 原则）编写。· 跟随 ML 和 MLOps 的方向保持步伐，
查看英文原文
This role can be based anywhere in Australia with remote first work or in Adelaide or Perth with Hybrid working where our offices are located.
Critical minerals are the bottleneck in the clean energy transition, and finding them faster is as much a machine learning problem as a geophysics one. At Fleet Space, our satellite-enabled exploration technology helps generate the data and the models to find deposits faster and more efficiently.
To help continue to build and scale our products we are looking for a Senior Machine Learning Engineer to help build out our ML operations to develop, mature and deploy the models and systems behind that work. You will partner with our another Senior ML Engineer & our Data science team to take their research models from notebooks and prototypes through to production-grade products, and you will own the pipelines, GPU environments and infrastructure those systems run on.
This is a role with a focus on scaling our ML ops alongside our Senior ML Engineer. You will design leakage-safe evaluation, verify that uncertainty calibration holds as new data lands, and keep every model's lineage reproducible from config through to prediction.
You will also own our ML experiment platform: MLflow tracking and the model registry, run identity and provenance conventions, and the environments and dependencies that training jobs run in. Where the platform team has shared infrastructure modules you will reuse them, and you will contribute new patterns back.
What You’ll Do
· Work with the data science team to transition research models into production-ready products, with code quality, scalability and best practice built in rather than retrofitted.
- Build and maintain robust, efficient ML and data pipelines and the infrastructure behind them, including configuring GPU environments for scalable ML systems.
- Own model quality across retrain cycles: leakage-safe evaluation, calibrated uncertainty that holds as new data lands, and reproducible lineage from config to prediction.
- Own the ML experiment platform end to end: MLflow tracking and model registry, run identity and provenance conventions, and the environments and dependencies training jobs run in.
- Partner with the platform team, reusing shared infrastructure modules where they exist and contributing new patterns back.
- Write clean, testable Python (SOLID principles) backed by unit tests and automated end-to-end tests.
- Keep pace with where ML and MLOps are heading, and bring the useful parts back into how we work.
What We Are LookingFor
- Deep expertise in machine learning across modeling, data engineering and architecture, with proven experience as a software engineer focused on ML
- Alignment to our stack; Python (Pandas, NumPy, scikit-learn, TensorFlow, PyTorch), MLflow, AWS, Databricks, Terraform, GitLab, and Zarr and Icechunk for big-array data
- ML Engineers that enjoy and want to focus on ML ops
- Experience with Geospatial data would be a big plus
- Experience in startup environments, building ML products and environments from early stages
- Experience across the full ML lifecycle, from data preprocessing and feature engineering through to training, evaluation and deployment.
- A track record of taking research artifacts (notebooks, scratch apps, prototype models) to production-grade solutions that scale, for example models served as a service.
- Data engineering from first principles: deterministic, idempotent pipelines, schema and contract discipline at data boundaries, and deliberate failure and retry semantics.
- Software delivery from first principles: continuous release with automated gates, observability, and clean design in application architecture.
- Solid grounding in MLOps: model development, deployment and data versioning, ideally in a startup or scale-up environment.
- Strong Python, with the common data science and ML libraries, and solid software engineering fundamentals across data structures, algorithms and design patterns.
- Hands-on AWS and Databricks experience, managing infrastructure as code with Terraform and designing to well-architected principles.
- Experience with code and data version control such as GitLab, MLflow and Icechunk.
- A high-ownership mindset: you balance technical debt against speed of value delivery and make that trade-off pragmatically.
- A high-agency mindset: you are comfortable where structure does not exist yet and needs to be built, and you treat iterative delivery as a means to that end rather than a ceremony.
- Exposure to big-array data such as remote sensing workflows, including Zarr and Icechunk, would be an advantage, as would experience with big data technologies, contributions to open-source ML projects, and comfort working with AI-assisted development workflows such as agentic coding tools and AI code review.
About Fleet Space
Fleet Space Technologies' vision from the beginning was to build technologies to help humanity explore and connect the Earth, Moon, and Mars.
This led us to develop and launch one of Australia's largest constellation of satellites, create our satellite-enabled mineral exploration technology (ExoSphere), and send Australia's first seismic technology (SPIDER) to the Moon in 2026. ExoSphere, our end-to-end mineral exploration solution powered by space and AI, aims to accelerate critical mineral discovery needed for the clean energy technologies foundational to humanity's future.
As part of Fleet's founding vision, we also apply our technology to innovative solutions for defense and space exploration, including an upcoming mission to the Moon. We are headquartered in Adelaide, South Australia, with growing teams in the USA, Canada, and beyond.
Company Culture
As a company, we are revolutionising the exploration of new worlds with advanced space technology to build a more prosperous future for humanity.
As a culture, we are ambitious, innovative, and collaborative and we put our customers first in everything that we do.
Our company values guide us to our North Star of ambitious, collaborative success: AD ASTRA (“To the Stars!”)
- Add Value: Obsessively add value for our customers
- Drive Excellence: Benchmark against the best
- Agile Action: Take Action. Independent, fast, frugal action
- Seek Truth: Be curious. Explore
- Take Responsibility: Make decisions and own them
- Radical Ideas: We are unique. We do things differently
- Always Deliver: We always find a way to get it done on time
Our Benefits Include
- We have an extremely flexible work culture, with a mix of onsite, hybrid and remote workers who take time for school runs, exercise and appointments. It's about getting the work done, not time at desk.
- Equity (ESOP) grants.
- 20 days of annual leave + 10 extra Wellness days per year.
- Learning budgets.
- Access to confidential Psychologist appointments via our Employee Assistance Program.
- Plenty of opportunity to be part of an amazing STEM program to create the next generation of explorers.
Equal Opportunity
Fleet is an Equal Opportunity Employer; employment with Fleet is governed on the basis of merit, competence and qualifications and will not be influenced in any manner by race, colour, religion, gender, national origin/ethnicity, age, sexual orientation, gender identity, marital status, mental or physical disability or any other legally protected status.Originally posted on Himalayas