
Image Classification
We worked with hundreds of thousands of photographs to classify visual evidence for compliance and fraud-detection workflows.
Founded in Sydney, Australia on August 2014, Matrix OS was created to make DevOps and Software Deployment Automation tools more user-friendly for software developers.
Between 2014 to 2017 we focused on Research & Development on distributed systems theory, type theory, category theory, machine learning, systems administration, network protocols, and programming language theory. We also created a website listing container orchestration tools: http://datacenteroperatingsystem.io/ after experimenting with Docker.
From 2016 to 2019 we formed a team to build Matrix OS, our core product. In 2019, we applied our experience building highly scalable infrastructure and machine learning applications through a period of consulting work. Today, Matrix AI focuses on technology at the intersection of artificial intelligence, cloud systems, and cybersecurity.
Matrix AI previously delivered applied machine learning and data infrastructure projects. These examples are retained as a record of the capabilities and experience that informed our later work.

We worked with hundreds of thousands of photographs to classify visual evidence for compliance and fraud-detection workflows.

We applied object detection to aerial imagery and other large image datasets to locate and count objects automatically.

We used segmentation to identify and measure material regions in mining photography, including estimates of visible gold grades.

We built containerized data-processing and machine-learning pipelines using distributed systems and cloud infrastructure.

We piloted an intensive Introduction to Deep Learning course with NetScout (NASDAQ: NTCT), training senior and principal engineers on machine learning with a focus on deep neural networks.
The 2.5-week program covered practical foundations including backpropagation, convolutional neural networks, recurrent neural networks, autoencoders, generative adversarial networks, and related deep-learning systems experience from our applied consulting work.
For general enquiries, partnerships, or questions about our work, reach out to us via email at enq...@matrix.ai.
We're always looking for new team members. See our current openings to get in touch and have a chat.