This ad will automatically close in seconds Close

Building the modern software factory

By Soumik Ghosh Dec 6th 2016
Building the modern software factory

Software makers are shifting to a factory production mode. Here’s what Otto Berkes, CTO, CA, believes can help make the cut.

The rise and rise of the app economy is making the enterprise sit up and shake off the slumber. Companies have to be quicker, more intuitive, and definitely more secure.

Otto Berkes, author of the book ‘Digitally Remastered’, and already a brand name in the digital transformation space, shares his two cents’ worth in a tête-à-tête with COMPUTERWORLD.

What is CA Technologies' rationale to stitch together app platforms with mainframes; mainframes being the company's long-standing cash cow?

Mainframes has been such an integral part of so many large enterprises for decades. We've got a very strong presence in the mainframe segment, and that continues in the enterprise.

When we look at the application economy, and customer-focused use cases enterprises are responding to, and if you look at the pipeline of technology all the way from an application running on a mobile device to what actually transpires in processing a request or surfacing some data that's needed by the application, you'll often find a mainframe as part of that chain.

The fact is that the mainframe has, and continues to do an incredibly good job at what it does. It's an incredible transaction processing engine, it's incredibly robust, it's secure, and it delivers real proven value. 

The fact is that the mainframe has, and continues to do an incredibly good job at what it does. It's an incredible transaction processing engine, it's incredibly robust, secure, and delivers real proven value. 

So, we'll be stitching an app on top of the mainframe layer, and to the extent that we can deliver a seamless end-to-end tools and technologies to be able to assemble an optimized pipeline that includes the mainframe.

You have emphasized the need for companies to build a modern software factory that steers away from the 'waterfall' development model. Could you elaborate on that?

When we look at what enterprises need through the lens of digital transformation, our goal is to provide the best in class tools. The goal at the end of it is to have an optimized software factory that can take an idea and deliver it as an outcome to the customer in the fastest amount of time, and at the highest level of quality possible.

The things we're providing, like test data management, or the security solutions, are ensured that only privileged users can access your infrastructure. We're making sure that you, as a customer, can build these software factories that are optimized for your business and customers.

Businesses are no longer content to wait for months to roll out a new software. What can companies do to enhance agility, and get into the software factory mode?

There're a number of factors. One is automation; reducing the chance for errors through optimization; the ability to release new code into production in small batches, at higher frequencies. 

Also, everything is tested and retested in a predictable manner. It really is about changing both the mindset and the methodology of how you build a software, design it, and release it into production. It's about creating a much more iterative feedback loop throughout the entire process.

How is CA Technologies harnessing embedded analytics and machine learning into its existing solutions?

We've made huge strides over the past year or so. One of the things we announced at CA World is a new analytics engine called 'Project Jarvis'. The goal of this project was to create an analytics engine that we could use to build new analytics capabilities into existing products, and also build new products that were focused on analytics.

Now, 'Project Jarvis' is the first CA accelerator incubation to go through the incubation pipeline and exit into a product organization.

Also, a dominant portion of the Jarvis development team is based here in Bangalore. So, it's a great example of innovation capabilities that exist at our development centers here.

Security is a killer feature. It's a sword that swings both ways. It can absolutely kill your business if you have serious security flaws. It can also be a fantastic enabler of your business if you can give your customers the confidence in your security capabilities.

In your book, 'Digitally Remastered', you touched upon the dire need to integrate security to fuel digital transformation. What can companies do to ensure that they have a robust security infrastructure in place?

The common approach has been to build security as the last step in the process of developing a new application or service.

What I advocate is that you need to integrate security deeply into the entire development pipeline - all the way from design up to deployment. Security is not something you add on after the development process, it's something that you absolutely need to build in.

I would expect developers to be thinking about security and applying security tools as an integral part of the development process, and not something that's applied right before release. 

Related: App Economy Changing CIO Role: Stephen Miles

Security is a killer feature. It's a sword that swings both ways. It can absolutely kill your business if you have serious security flaws. It can also be a fantastic enabler of your business if you can give your customers the confidence in your security capabilities.

Also, another important factor to bear in mind: Implementing security is enabling user experience, not getting in the way of it.

What in your experience, are the pain points companies might encounter when they deploy security at the end of the digital deployment?

What happens when security is deployed at the end stages is a classic case of "Oops! We didn't think of that." So, you'll have to go back, all the way upstream - sometimes to the architectural level, and re-implement entire chunks of functionalities. 

The security flaws that you discover at the end of a process are much more expensive to fix than flaws that you've identified earlier. 

If you treat security as an afterthought, you're not going to necessarily have the most robust offering. But even more important is the customer experience. You want to make security inherent in the product that you're offering, so it seems very natural, and not a deterrent. 

CA Technologies has deployed a payment security solution that relies on data science and analytics. What makes this a game changer?

Really, the whole point of payment security is to ensure the integrity of transactions without getting in the way of it. That includes not having to ask for any additional security information from the customer, by harnessing data science and analytics instead.

We've invested a lot of time and effort into the data science aspect of our solution to be best in class. At the heart of the solution is the intelligence behind it. The algorithms and data science that create an intelligent view about a transaction or a pending transaction based upon a large number of factors that through the solution that we provide, create a very robust picture of that transaction.

We've managed to strike the right balance between securing transactions and minimizing friction. And this solution has been predominantly developed in India.

Companies can get caught up in the "vendor soup", as you like to call it, when they partner with multiple vendors offering multiple solutions. What do you think is the work-around?

Sometimes there's a tendency to throw lots of solutions at a problem, and I think that the right approach comes from taking the long view. It's important to have an architecture that makes it very clear as to what components you need, what point in time you apply them, and what your desired outcomes are.

So, multiple vendors is not a bad thing. But, you need to have a clear understanding of the integration points, and that comes back to having a very clear view of the architecture you want to build. 

Ultimately, you want to build something that works seamlessly as a whole, and that requires some real architectural discipline and forethought. 

We see the emergence of containers in the container ecosystem, so we're making investments to ensure that we provide container support and functionality to our customers based on their growing adoption of containers. 

What are the company's prime focus areas in 2017?

We made some great headway in analytics and integrating analytics into existing products. i expect us to continue on that path. Analytics is going to be an absolute core path for our business.

We see the emergence of containers in the container ecosystem, so we're making investments to ensure that we provide container support and functionality to our customers based on their growing adoption of containers. 

We're also going to continue to invest in our organic incubation efforts, so we're going to continue moving forward with our CA Accelerator program that we launched this year. 

And that's not just in CA Accelerator, it's across the board. If you look at something like advanced experience analytics, that was an organically developed product. So, you could expect to see more organically developed products in the next year.

LATEST interviews