From medical image analysis to manufacturing and construction protocols, Datature develops computer vision models that typically require specialized technical expertise and technology. To remain sustainable, the company needed a platform that could scale with its growing datasets, particularly as it began to add new microservices to its platform.

With Google Cloud as its core infrastructure, including Cloud Storage and Google Kubernetes Engine, Datature enables customers to build generic pipelines for data preprocessing, model training, and deployment with just a few clicks.

Through seamless access to GPU capacity that can be packaged for individual customer use cases, provision of global infrastructure to support business growth, and secure, scalable storage for image and video datasets, Google Cloud enables Datature to lower the barrier to computer vision for organizations across medical, manufacturing, retail, utilities, and other industries.

The organization worked closely with Google Cloud Partner Searce on the deployment. Searce provided support by debugging issues, and, with a deep understanding of Google Cloud services, crafting an architecture that solved the challenges posed by large clients with complex requirements.

Using its Nexus platform and Google Cloud GPU high performance processing capacity to accelerate model training, Datature was able to support a university research team to create a computer vision-powered brain lesion detection system. The versatility of the platform is demonstrated by a manufacturing company's use of the same pipeline and model architecture to create a computer vision model that helped check products for defects.