Chris Buckley

Chris is never satisfied with just achieving success, he is constantly looking for the next edge and is always focused on business outcomes, in a commercial context. Having solved perplexing problems within the most complex global environments for nearly two decades.

Since joining us at Virtual Clarity he has developed distributed systems platform strategy, and served as lead architect for global enterprise private cloud.

What you might not expect is that he combined being a part-time racing driver with being the founding co-chair of the Open Data Center Alliance’s infrastructure workgroup (he still races by the way). Chris is guided by a passion for identifying the right problems to solve from a business perspective, feel free to challenge him.

Connect with Chris.

View articles by Chris

Transforming digital-ish assets on AWS- Using AWS machine learning services to transform core applications

An argument one often hears made about cloud migration is that, if you can just get your applications and workloads over to the platform, transforming them to use native cloud services will be simpler than trying to do it on the way in.

It’s an argument with some merit but it’s always been hard to miss the self-serving aspect when looked at from the cloud providers’ perspective. For the argument to really work, providers need services on the platform that make the transformation part of the story much easier than it would be otherwise. In the impressive laundry list of machine learning services announced by AWS at Re:Invent, there was at least one service that might play that role.

At Virtual Clarity, we regularly help our customers build business cases for transformation that consider the real costs and benefits of different kinds of migration and transformation. We offer customers both the big picture (transformation of an estate) as well as detailed examinations of individual, often critical applications. We have worked with several customers to look at the applications that run their business to assess how they might run in AWS.

We have seen a recurring pattern with these applications - that they rely on what might be called digital-ish assets. Examples of digita-lish assets abound in industries like mortgage services, insurance and healthcare. These are the forms and deeds that support historic mortgages; they are the policy documents and assessor notes generated in insurance, as well as those notes your doctor wrote that even they could not really read. All those assets have been scanned to be held in document repositories and workflow systems that link assets to structured data in often ancient applications trapped on mainframes, mid-range and traditional open systems platforms.

When we look at the transformation opportunities for applications of this kind, what we repeatedly come up against is the cost of transforming the albatross of digital-ish data. Without bringing this data into a transformed application, businesses are stuck maintaining traditional applications or old archives alongside new. Moving applications with digital-ish assets into a public cloud platform for transformation is a difficult sell when the only realistic mechanism to transform the data was to use human beings to perform traditional data entry tasks to transcribe data out of digital-ish assets (footnote: there are 3rd party products out there that apply machine learning and other techniques to this problem but they can be difficult to experiment with and require setup care and feeding that could make the business case tough).

All this might be changing, thanks to the packaged analytics services that were announced at AWS Re:Invent. For example, the Textract service looks like it might make the migrate and transform pattern viable for this class of core business applications. Textract applies packaged machine learning to extract meaningful data from digital-ish assets, creating structured information that a transformed application can really work with. Not only does this offer the possibility of eliminating the legacy systems that employed digital-ish assets, but it opens up new options for businesses to use the data trapped in digital-ish assets - to apply analytics to find customers who might benefit from new products, for example. Importantly, Textract has commoditised pricing and is offered as a service.

The slew of machine learning announcements at Re:Invent was the big surprise of the event for me. I was expecting a machine learning story, but nothing like the sheer range of packaged analytics products and services that together could make using machine learning a dramatically simpler proposition. Assuming these services do what they say on the tin, they offer a dramatic increase in the productivity of data scientists and could open up whole new classes of business and products to begin to apply machine learning. Although Google Cloud Platform and Azure, for example, have had packaged machine learning solutions for a while (e.g. GCP’s Cloud Natural Language API, Azure’s Content Moderator, etc.), the richness of the AWS platform coupled with the announcements at Re:Invent perhaps herald the great democratisation of machine learning.

With packaged analytics services like Textract, AWS are setting out their stall for changing the calculus businesses apply to application transformation decisions. Expect to see some big transformations of complex core applications in the coming year.