CSIRO’s Australian e-Health Research Centre (AEHRC) has led the way by becoming the first public sector organisation in the world to publish a machine learning-based health product on the AWS Marketplace.
“CSIRO is the first public sector agency globally to offer a genomic analysis research product through AWS Marketplace,” says Iain Rouse, AWS Public Sector Country Manager. “We welcome CSIRO as a public sector pioneer on AWS marketplace.”
Freely available during the promotional period, CSIRO’s analytics platform VariantSpark enables researchers to pinpoint the genes that cause disease in large human genome case-control cohorts. With an interactive user interface and by automatically managing several hundred CPUs in the cloud, VariantSpark aims to democratize machine learning based genomic research.
CSIRO’s Dr Denis Bauer, who leads the team that developed VariantSpark, says making software globally available through digital marketplaces is an essential next step as “Data analytics as a Service” business models take off.
“Digital marketplaces can help people find the right specialized tools internationally that are also capable of scaling with their business needs without being slowed down by individual licence-negotiations,” Dr Bauer says.
“Through a digital marketplace we can make VariantSpark available on a powerful compute cluster and deploy it in the customer’s account. Taking the analytics to the customer data is the ideal solution for the health space, as it maintains the privacy and security of the data, and reduces unnecessary data transport costs,” Dr Bayat says, the lead researcher in the project.
Increasing access to academic-grade analytics to support industry data science quality
Brendan Hosking, lead developer for the product, says global digital Marketplaces help host, socialise and ultimately deploy algorithms that would have otherwise been locked away in academic papers or be limited by local market demands.
“The guidance of AWS’s world-class expertise during the onboarding process has helped us develop an industry-grade security-hardened software, a standard substantially harder to reach through traditional commercialization processes ,” Hosking says.
Dr Bauer says digital marketplaces can help deliver demand-ready services more easily.
“Achieving a predictable, renewable revenue stream that scales to more customers without additional investment provides the necessary budget for ongoing research and development into continued productivity – or in our context, helping deliver public-good research,” Dr Bauer says.
“Offering advanced research analysis tools to industry could raise the standard of data analytics, leading to more effective, data-driven business decisions, says Brendan Speet who coordinates CSIRO’s cloud enablement
“Enlisting the IT industry to help develop analysis software that’s more shareable and reproducible can hopefully accelerate research outcomes, especially in crucial areas like human health.”
Getting digital marketplace-ready
Digital marketplaces are not new, with AWS Marketplace launching in 2012 and the Azure Marketplace existing even earlier. But the general rise in mainstream cloud-usage – with 90 per cent of organisations using some type of cloud service – has led to the recent increase in digital marketplace adoption.
“The Marketplace reduces the risk of launching an early product as it enables companies to leverage established distribution channels and removes geographical barriers”, says Janet Fox, business development manager at CSIRO’s AEHRC.
Despite these benefits, a recent Deloitte report found offering products through the digital marketplace can be difficult. To be successful on the Marketplace, products need to be centred around customers and remain under continuous development, offering new features and expanding its customer base, which is a challenge under the traditional product-centric business model.
With years of experience developing customer-centric, agile, cloud-based solutions, AEHRC was well-placed to move their solutions into the digital marketplace.
Dr Bauer says the scientific method her team is trained in is perfect for the digital marketplace model.
“We continuously scope improvements in the algorithms in response to emerging evidence, whether that’s experimental results, feedback from customer engagement or advancements demonstrated by the community,” Dr Bauer says.
“We’re keen to be able to bring these skills to our future R&D projects and collaborations, creating end products suitable for use on digital marketplaces.”
To talk to Denis and the team about collaboration, contact email@example.com.