My client is Australia’s leading AI startup - they are experiencing hypergrowth and building true foundation models for the geospatial domain.

While most of the industry is still focused on text and images, they’re training multi-modal spatial foundation models that understand the physical world at scale. These models work with point clouds, lidar, geospatial data, and complex asset information to solve problems that standard computer vision and NLP approaches can’t touch. The product is solving these problems at scale, the data is messy, and the performance and deployment challenges are genuinely new.

They’re growing at 2x YoY, backed by a sharp, high-ownership team that cares about shipping real systems, not demos. They need a Senior MLOps Engineer who can keep the ML training and serving infrastructure reliable while the ML research and applied teams moves fast.

What You’ll Do

- Own the end-to-end ML pipelines that take geospatial foundation models from data ingestion through training, evaluation, and production deployment
- Build the automation layer — CI/CD for models, artefact registries, reproducible environments, and infrastructure as code — so experiments become deployed services without manual effort
- Keep models healthy in production with monitoring, drift detection, data quality checks, and clear ownership when things break
- Manage distributed training across GPU clusters (cloud, on-prem, and neocloud), focusing on utilisation, reliability, and fast debugging




- Deploy and scale inference services that can handle variable load, with attention to latency, cost, and deployment constraints
- Improve the day-to-day tooling and standards the ML team relies on as we grow

Key Skills Needed

- Hands-on experience building and running ML training pipelines, model serving, and production monitoring
- Strong Python and enough PyTorch (or equivalent) to debug training jobs and understand what’s happening under the hood
- Practical experience with distributed training and GPU infrastructure - scheduling, resource management, and diagnosing performance or memory issues
- Cloud experience (AWS, GCP, or Azure), Kubernetes, Docker, and infrastructure as code
- Experience with model monitoring, data quality frameworks, and preparing real-world datasets for ML
- Positive engineering judgement — you write clean, maintainable code and know the difference between a temporary fix and a proper solution

Bonus: CUDA/kernel-level work, exposure to point cloud or geospatial data, or experience deploying into constrained or air-gapped environments. - These are nice-to-haves but not essential. What’s In It For You

- Top of market salary and meaningful equity
- A smart, high-calibre team working on real foundation model research with production impact
- The chance to help define how geospatial AI is built and deployed at scale, inside Australia’s fastest-growing AI company

If you want to work on true AI/ML problems with a team that’s actually shipping, we’d like to hear from you.

📌 Senior MLOps Engineer - $1bil AI Startup - Foundation Models (Australia)
🏢 TENAX TALENT
📍 Australia