Blog icon

Mr Ben Leighton

Engineering Lead l Environmental Water Science Platforms, AI, Simulation, ML, Geospatial

https://people.csiro.au/l/b/ben-leighton

Contact

Please enter a valid first name

Please enter a valid last name

Please enter a valid email address

Please enter your message to send to the contact

Biography

Lead Delivery Engineer | Senior Platform, AI, & Data Architect

Ben Leighton bridges the gap between scientific research and operational engineering, delivering high-performance systems that integrate leadership, software engineering, data science, and AI/ML. Driven by a commitment to equitable and sustainable science, Ben translates complex research into robust, production-grade infrastructure.

At CSIRO, Ben architects, builds, and leads the development of innovative solutions for scientific AI, knowledge engines, and large-scale data pipelines. He combines deep technical expertise in Climate Modelling, Water Science, AI, geospatial systems, cloud infrastructure and machine learning with a proven ability to lead diverse teams, ensuring that cutting-edge research is successfully operationalized for real-world impact.

Key Achievements & Technical Delivery:

  • BlueLink / Next Generation Coastal Modelling (NGLP): Ben is the lead discipline engineer in BlueLink coastal systems uplifting research workflows into critical secure operational systems. He is an expert in engineering of complex production workflows for ocean modelling. He led a team structuring petabyte-scale national ocean data climate modelling products, utilizing H3 geospatial indexing for optimized spatial partitioning and orchestrating high-performance pipelines using Apache Airflow, Xarray, and Dask.

  • National Bushfire Intelligence Capability (NBIC): As Lead Delivery Engineer, Ben pioneered the NBIC data delivery platform (including the Aptly and CogBoard systems). He architected a secure, serverless geospatial delivery engine on AWS, utilizing Cloud Optimized Raster (COG) to scale performance and sophisticated IAM policy management for security. The system supports over 100 stakeholders across Australia with real-time, high-resolution climate futures hazard data.

  • Water Science, Generative AI & Knowledge Engines: Most recently, Ben delivered a Geospatial-enabled RAG (Retrieval-Augmented Generation) AI system for the North Australian Water Resources Assessment (NAWRA) and Murray Darling Basin. Leveraging LangChain, OpenAI, and FAISS vector stores, he built a system that allows external stakeholders to intuitively query and extract context from complex scientific reports and geospatial data. He is engineering systems to use AI Agentic approaches to integrate the Geofabric Hydrological Layers for decision support. He has extensive experience developing Water Science modelling systems including AWRA and eWater Source Catchment Modelling products.

  • Hybrid AI & Surrogate Modelling: Ben implements hybrid physics-ML techniques, creating surrogate models that accelerate scientific simulations. Using Scikit-learn and Keras, he has helped develop novel approaches for coastal ocean modelling and predictive analysis. Ben applies deep expertise in ML lifecycle management, solving complex problems in feature engineering and bias reduction to ensure rigorous scientific validity.

Community & Strategic Leadership:

  • Community of Practice: A dedicated technical leader, Ben leads CSIRO’s Tuesday Morning AI community, a thriving cross-disciplinary group of experts driving the adoption of Generative AI and cutting-edge machine learning practices across the organization.

  • Project Coordination: Ben is a certified SCRUM Master with a track record of coordinating distributed teams for delivery of production systems across CSIRO and partner organizations (e.g., BOM, AWRA, AURIN).

Skills: Science Terrestrial water, ocean, fire science, data science, deep learning, surrogate modelling, hydrological data infrastructures, keras, sklearn. Web Full stack development, OpenLayers, MapBox, FastAPI. Data Engineering Dask, Xarray, Spatial DuckDB, H3 and spatial indexes, SLURM, Airflow. Cloud AWS CDK, Terraform, GCP CloudRun, ABAC, Auth0, Secure Streamed DataLake. AI FastMCP, LangChain, AI-first engineering, LLM uplift and training, community building, LLM batch workflows, FAISS vector stores, Unstructured document parsing, RAG, Agentic Architecture

Career History & Impact:

Ben’s portfolio spans a diverse range of impactful projects. On the Energy Use Data Model (EUDM/NEAR) project, he developed cloud-based time-series processing pipelines utilizing AWS RDS for scalable data management. As a software engineer for AURIN, he was an early adopter of Docker-based approaches, pioneering reproducible data analysis deployments. He has also played key roles in the Bioregional Assessments metadata systems and the Australian Water Resources Assessment (AWRA), bridging the gap between software developers and scientists to deliver national water intelligence. He was an engineer on the eWater Source platform.

Ben believes that value is maximized when diverse people, skills, and communities integrate. He is committed to fostering collaboration and designing engineering solutions that enable a better, more sustainable future.

Current Roles

  • AI architect and engineer
    Enhanced RAG Knowledge Systems, Agentic Systems, Citation Grounding

  • Full Stack Software Engineer
    Python, HPC, Cloud (AWS, GCP), Web

  • Data Engineer
    Cloud and On Premises Spatial, Temporal and Modelling, Data Pipelines

  • Technical Leadership
    Designing, Developing, Communicating and Coordinating, Next Generation AI and Data Platforms

  • Client Liason
    Business Analysis and Development, and Contracts Coordination

  • Environmental Data Analyst
    Spatial Data Assessments and Algorithms

  • Software Engineer
    Geospatial Data Lakes, Data Visualisation

Academic Qualifications

  • 2000

    BA (Philosophy)
    Australian National University

  • 2000

    BSc (Biology)
    Australian National University

  • 2015

    Machine Learning
    Coursera

Professional Experiences

  • 2001-2003

    Software Engineer
    Wizard Information Systems

  • 2003-Current

    Software Engineer, Data Scientist, Data Engineer
    CSIRO

Achievements and Awards

  • 2016-2016

    CSIRO Medal for Impact from Science
    CSIRO

  • 2012-2012

    Excellence in Support
    CSIRO Land and Water

  • 2007-2007

    Partnership Excellence
    CSIRO Land and Water

  • 2017-2018

    CSIRO Collaboration Award
    CSIRO

Published data & software

We have published items by Mr Ben Leighton