Author: Mark Hillary

The business journals used to say that outsourcing and offshoring was replacing the need for companies to hire people. Now it’s the robots that are supposed to be taking our jobs. Business journals and commentators across the world are suggesting that a wave of automation driven by smart Artificial Intelligence (AI) systems will largely replace the need for workers.

Enterprise technology used to be completely unrelated to any of the systems we might use on a computer at home – think of ERP or CRM systems. Who would be building a CRM system at home, yet it’s a common tool in the office. However, now that many enterprise tools are delivered via the cloud using a web browser it feels remarkably similar to be using tools and systems at home or in the office.

So is the Internet of Things (IoT) any different? The IoT has largely been talked of as an automation tool to assist in the home – the smart home where every device is connected and controllable. But how does the IoT work in the enterprise environment and should it be approached differently to the home?

There are in fact many differences that enterprise IT managers need to be aware of and this interesting article lists ten talking points. I’m going to just focus on what I consider to be the most important three here:

1. Security; you might be comfortable capturing personal data, such as your heart rate, on devices at home and sharing that information with your own personal cloud, but in the enterprise environment systems handling highly personal information need strong protection – just assuming a password is enough is negligent.

2. Reliability; enterprise and industrial IoT systems may be deployed far from your headquarters in remote and hostile environments. You need to know that a sensor will have a lifespan that can be measured in years, not weeks.

3. Programmability; IoT systems produced for personal use at home are largely out-of-the-box solutions. Even systems that can be configured in many ways, such as the Amazon Echo used for controlling home devices, are generally used in default mode most of the time. In the enterprise environment it is critical that your IoT devices and general environment can be changed to suit your needs – the default setting is unlikely to meet your needs.

This is an interesting paradox. As we have observed consumer and enterprise applications becoming more and more similar, the world of IoT remains distinct. The CIO, or general IT management, need to create an enterprise IoT strategy that moves beyond just installing systems that would be more useful in a smart home environment.

Author: Mark Hillary

Most people working in IT today know about the cloud and how cloud-based systems can offer immediate access to storage or computing power easily. Many companies now use a cloud strategy to ensure they can ramp up and down on available systems or storage – it’s a common theme of discussion when planning an IT strategy.

But what is a Cloud Management Platform and in particular how can one be useful to service providers offering IT services to their clients?

Cloud Management Platform (CMP) is a term coined by the industry analyst Gartner. The analyst firm wanted a way to describe products or tools that help companies to optimize and manage their cloud infrastructure for cost, security, and operations. A good CMP strategy should allow users to maintain control over dynamic and scalable cloud environments.

So it is really just a control system that allows the user to maintain dynamic control over their cloud system. Instead of frantically calling a service provider and asking them to quickly scale up storage capacity in their contracted cloud, the customer should be able to use a CMP to just scale their cloud – it should be as easy as sliding a bar on a control panel.

The major CMPs on the market today will all offer these different aspects of cloud management:

• Cost
• Security
• Performance

These are the three main areas where the customer can manage their cloud and cover more specific areas such as budgeting, rightsizing the cloud, compliance and monitoring, and creating alerts and other analytics.

CMPs are still quite new tools and so there are still different types of tool available. Some are very focused on specific issues, such as controlling security risk or optimising costs. Some allow the option to manage multiple clouds, so if you are using both Microsoft Azure and Amazon AWS then you can manage both clouds from a single tool. Some CMPs even allow the tool to manage a cloud and systems you have on site simultaneously.

Clearly this is an emerging area. There are many new tools solving problems that have only recently become apparent. It was only a few years back that companies started taking cloud-based services seriously and it has become clear that it can be difficult to control the various aspects of a cloud-based system – such as cost control and security. In addition, if a service provider is delivering services on a cloud that is owned and managed by the client then the client needs an easy way to manage areas such as security in partnership with their service provider.

CMPs are still new and it is therefore difficult to advise on which one is perfect for each client, but there is a clear need to work on CMP selection with your service provider as any chosen solution must work for both client and service provider. I’ll explore this question next time here on the IBA Group blog.

You can also read about the IBA Cloud Platform

Author: Mark Hillary

I have written about DevOps recently on this blog, but to summarise the concept, DevOps refers to Development Operations. It is a combined set of software development practices that bring together the development of software with IT operations. The aim is to improve the systems development environment so the software lifecycle can be shorter – it’s bringing software development closer to the business that it serves.

So far it sounds like DevOps is just focused on software development and the environment used to build software systems – it’s all about coding and process. So what are the skills needed to work in DevOps? Logically you might assume that ‘hard’ subjects such as science, technology, engineering, and mathematics (STEM) might dominate, but I read an article recently that turns this expectation upside down.

Tech Beacon magazine listed four key ‘soft’ skills that everyone working in DevOps requires and none of them are focused on STEM skills. They are:

1. Collaboration and Communication
2. Empathy
3. Customer Experience
4. Problem Solving

Why would these be the essential skills and not coding or process design? Well think for a moment about what I said in the introduction – we are bringing the software development process closer to the business that needs this technology system. So some of the key skills will be focused on that process of getting closer to the sponsoring business people.

Collaboration with people outside the IT team will be essential and the ability to communicate technical problems to non-IT professionals. Empathy implies more listening – especially listening to the people who want the system built for their business. Trying to put yourself in the shoes of the customer so you can improve the customer experience is also an important skill that many people ignore, and the ability to solve problems as they are thrown at you is extremely valuable in any team.

I would argue that these four skills are essential for any DevOps team. If you can find people to join the team with all these skills then it is almost certain they can learn the technical skills you need them to use. If you hire for technical ability only then it will be much harder to create great communicators or problem-solvers through training.

It may seem like the opposite to conventional wisdom, but sometimes the best team members in a technical DevOps team are the least technical.

Author: Mark Hillary

The industry analyst firm Everest Group recently reported their latest research into the development of the Robotic Process Automation (RPA) market and they have some interesting findings. The market is getting larger and more competitive, but the services provided are also increasing in complexity too.

“The market for RPA technology is becoming more complex, with highly competitive and evolving product offerings,” said Sarah Burnett, executive vice president and distinguished analyst at Everest Group. “The Everest Group Products PEAK Matrix is an unbiased assessment that uncovers vendor and product differentiators to identify the leaders in RPA technology based on research that goes deep into the vendors performance and product details, features, functionalities and more.”

The PEAK Matrix reporting method summarises the 22 companies studied along two different axes. The first is vision and quality, describing the ability of the company to successfully deliver the products they promise. The second is the market impact, how is their sales performance and impact on the wider RPA marketplace?

Everest Group classified the main RPA technology vendors into three categories of Leaders, Major Contenders, and Aspirants:

• Leaders: Automation Anywhere, Blue Prism, NICE and UiPath
• Major Contenders: Another Monday, AntWorks, EdgeVerve, HelpSystems, Jacada, Jidoka, Kofax, Kryon, Pegasystems, Servicetrace, Softomotive, Thoughtonomy, and WorkFusion
• Aspirants: AutomationEdge, Datamatics, Intellibot, Nintex, and Nividous

AntWorks, Automation Anywhere, Datamatics, Softomotive, and UiPath demonstrated the strongest year-over-year movement on both market impact and vision-and-capability dimensions and emerged as 2019 RPA Market Star Performers. WorkFusion scored as high, or higher sometimes, on the vision and quality measurement as the leaders, but it was felt that they have not quite made the market impact to be categorised as a leader – yet.

Other interesting takeaways from the Everest research include:

• Automation Anywhere, Blue Prism, and UiPath are the top vendors in terms of RPA license revenue, closely followed by NICE. Pegasystems leads in terms of revenue from attended RPA (RDA) licenses
• Softomotive has the highest number of RPA clients in the market, most of which are small-sized enterprises and SMBs. Witnessing almost 300% year-over-year growth in its number of clients, UiPath holds the second spot in terms of number of RPA clients
• Automation Anywhere leads in North America, which is the largest RPA market, and Latin America. Blue Prism leads in the UK and MEA markets, while UiPath leads in Continental Europe and Asia Pacific
• UiPath holds the highest market share by license revenue across horizontal functions such as F&A, procurement, and HR, while Blue Prism leads in banking and insurance industry-specific process areas. Pegasystems has the highest market share in the contact center space
• Leaders have moved away from perpetual licensing to annual/monthly subscription-based licensing models. Advances in RPA technologies and increasing client maturity are fuelling the rise of more output-oriented pricing models such as flexible usage-based (e.g., per minute/hour) and per-process or transaction-based models
• RPA solutions continue to evolve with a host of capabilities, such as computer vision, workflow orchestration, intelligent workload balancing, auto-scalability, and predictive SLA monitoring, to help enterprises achieve strategic business outcomes
• Attended RPA / RDA continues to witness increased demand in the market. AI-based next-best-action recommendation and interactive UI for on-screen agent guidance, which enhance worker productivity and help improve customer experience, are among the key RDA differentiators

This new research from Everest Group is comprehensive and insightful and was only just published in July 2019 so the information is current. I recommend browsing their findings because it is clear that RPA has moved beyond the hype cycle now and is becoming a serious strategy for many company executives.

Author: Mark Hillary

Last year the analyst firm HfS Research published a cautionary article on Robotic Process Automation (RPA). They warned that the technology and outsourcing market was hyping RPA beyond what it could possibly achieve – the talk of RPA changing the world is all hype.

This should be qualified by the more nuanced comments published by HfS. They stated that the RPA boom has been hyped, but not the more general focus on business automation. So the blogs and business journals that keep on breathlessly saying that bots will transform companies need to examine just exactly what they are saying. RPA cannot exist without a wider transformation of how a company functions.

Companies today need real-time data that converges across business functions. You cannot have a marketing database that is not connected to data your customer service team is using. You cannot drive business value today without your business data being real-time and accessible from across the entire business.

The hype around bots is that all this is automatic; but if your company still relies on paper documents then how are the bots going to process that data? To create a truly automated organisation requires more thought and planning than just the deployment of a bot in the back-office.

As more processes are digitised, more value can be created. More insights can be found. More opportunities to set bots searching for trends are created. HfS has actually been saying this for a long time and they have a digital office solution called OneOffice that broadly defines how organisations need to plan operations – if they want to use the data flowing through the company.

The real message is that RPA itself is not hype. There are many excellent examples of RPA deployments that have achieved genuine efficiencies for the companies using it – including customers of IBA. However, the idea that RPA is a silver bullet that can automate your business processes, making your team super-efficient overnight, is certainly hype. RPA is just one part of a transformation to a digital business environment. It is one tool in a complete arsenal of change related to the way that data is used.

We need to see a convergence of data analytics, cognitive solutions, and RPA. Big Data is another term that is often tossed around with very little understanding, but it needs to be understood by leaders. Big Data and data convergence are not just referring to smarter ways of using data. We are talking about the complete digital transformation of organisations so that data becomes the most valuable asset.

Many people have been shocked to recently discover just how much data Facebook has on their personal likes, dislikes, and preferences. It shouldn’t be a surprise. Facebook has offered a free service for years in return for data. That data allows them to know their customers inside-out, creating advertising opportunities that the traditional advertising industry cannot match. Facebook has redefined how advertising works and major advertising companies are struggling to keep up.

Whatever your business, you need to think about how data flows from customers to you and how it moves inside the organisation. How can you leverage this knowledge to create new opportunities – even a new business? RPA is just one ingredient and can help to automate some processes, but thinking about RPA should really just be your first step towards a complete transformation of how you are using data in your company today.

Author: Mark Hillary

Think about the consumer technology that you regularly use today. You probably have a smart phone, maybe a Kindle or other e-reader, maybe an Apple Watch or similar device that can access information from your phone. Maybe your car can hook up to your phone to offer in-car information. Maybe you have an Amazon Echo at home so you can access the Internet just by speaking?

All these consumer devices are available today and are accepted as normal. Most consumers expect to have a device that gives them 24/7 access to all the services and information that the Internet can offer.

So why isn’t enterprise technology like this? Many companies still issue phones that are not even smart and laptops that are too heavy to really be portable. The concepts of cloud computing and app store flexibility remain conceptual in many organisations. Why?

The obvious answer is that consumers have far less to invest than large companies. When purchasing technology, a CIO needs to set the agenda for several years. If things change during that time it can be difficult to shift direction or to keep up with the change. Individuals don’t face this problem.

This has led to the popularity of Bring-Your-Own-Device (BYOD) policies in many companies, where employees are offering cash to use their own equipment instead of what the company can supply.

But a small change in the strategic mindset can also have a major benefit to the enterprise. Commissioning new software solutions as apps, rather than desktop tools can encourage the workforce to be mobile. This can even encourage companies to create entirely new solutions for customers.

An app developed by IBA for use by a bank in South Africa allows bank employees to sign up new customers on the move. They can photograph the customer using their phone and capture details which are then shared with the central system of the bank – no forms, no waiting for an appointment. The new customers, the mobile bank employees, and the bank executives all benefit from the app approach.

It used to be that enterprise technology was years ahead of what people had at home, but now the reverse is true. It’s time for more company executives to take inspiration from the tools they use everyday – how can we use mobile devices and other common personal technology to create better business solutions for our customers?

Author: Mark Hillary

I mentioned that Big Data has progressed far beyond just being a business buzzword. There are entire industries being shaken to their core because a leading player finds a way to analyse their customers better than their rivals. Far from being a management trend, this is a strategy that will fundamentally change many industries.

But have you ever stopped to appreciate just how much data we are creating today?

Some excellent analysis in Business Insider recently explored this question. The problem is that people and companies are just creating so much data – it is increasing at an exponential rate. At the present rate we are doubling the amount of digital data that exists every two years.

But even though we are creating and storing these enormous amounts of data, only 0.5% of it is being analysed. There is so much data out there that companies, governments, and individuals feel swamped, unable to gain insights from it.

The Business Insider article features a comment that cuts to the heart of the Big Data issue: “You have to start with a question and not with the data,” says Andreas Weigend, former Chief Scientist of Amazon, now director of the Social Data Lab and lecturer at UC Berkeley.
Businesses need to start thinking about the insights they could get from their customers, to ask more ‘what-if’ questions. There are solutions out there in the data, but it is impossible to analyse every byte of data.

A typical analogy for the average person might be the difference between email and Twitter. You check every email, even if it is only long enough to decide that it should be deleted. However, you only check Twitter messages that are arriving as you are looking at the news stream, or you use intelligent filters and analysis to ensure that interesting messages are made visible.

Businesses need to start thinking of their Big Data strategy in the same way. How can insights be drawn out from the data they already have?

Author: Mark Hillary

I wrote recently on this blog about the use of SAP to create Digital Twins, a digital representation of a real system so it can be more easily monitored and controlled. As I mentioned earlier, this has been a common practice in aviation for several years. Engine manufacturers will always maintain a digital version of every engine they sell and ensure that the digital version is updated in real-time using sensors on the real engine.

In aviation the advantage of doing this is obvious – it allows for more efficient maintenance and safety procedures when the digital engine allows engineers to monitor real engines remotely. But I believe that the launch of SAP’s Leonardo system last year will really start accelerating the use of Digital Twins as a common business strategy.

It is a combination of technologies and strategies that are creating this possibility, but the three important ones are:

1. The Internet of Things (IoT); as the real world fills with sensors and connectivity as standard for almost every electronic device we will reach a point where systems such as a Digital Twin are essential just to stay on top of what is connected and what information it is reporting. In the home, this may only be devices such as a Kindle, iPad, Echo, phone, lightbulbs, or heating thermostat, but in the industrial environment it can easily be more complex and difficult to control.

2. Artificial Intelligence (AI); with so much data being created constantly by sensors we will need to apply AI principles to the data just to make sense of it all. For example, if your home thermostat detects patterns in the way that you prefer your home to be heated then it should be able to anticipate what you want before you change the settings.

3. Machine Learning (ML); the ability to look at every action and outcome by every sensor inside a network will allow the system to learn about the ideal outcomes and then to suggest recommendations in future based on earlier learning.

It is really the IoT that is at the heart of this development. Imagine the complexity of a modern industrial facility – a large brewery or car factory for example. Across the entire property will be doors, windows, pumps, and various robots that all need to be coordinated. Most companies with these facilities will already have some sort of control mechanism, but the Digital Twin makes an assumption that every component (pump, door, assembly-line robot) has in-built sensors. By taking a feed from all of these sensors we can build a complete virtual mirror of the plant.

The IoT facilitates this by ensuring that the real-time sensor data is available, then the AI system goes to work on spotting potential problems or just process flows that are unusual and alerting workers to places they need to check.

I have seen this type of system deployed for an office management system where every light, heater, door, and window is modelled in the system. I believe that we will see the Digital Twin concept growing much faster as companies find that they can create enormous efficiencies by improving what they do and spotting problems before they happen.

As Forbes magazine recently suggested, it will soon be impossible to plan any kind of digital transformation for your business without creating a digital twin first. The processes will be simply too complex for any one manager to understand from start to end. Not only do you need to map out all the existing components, you need to apply AI to oversee how the entire system is working.

Without these deep insights into the way your business functions at present any transformation plans will be impossible. Digital Twins are not just for those obsessed with being able to manage their existing IoT infrastructure, they are becoming an essential tool for managers who want to see how the future of their business might look.

Author: Mark Hillary

What is DevOps? If you don’t directly work on the development of IT systems then this might be a strange concept, but for any executive who needs to purchase new technology systems, or modify existing ones, this is an important concept to be aware of.

In short, DevOps refers to Development Operations. It is a combined set of software development practices that bring together the development of software with IT operations. The aim is to create an environment where the systems development lifecycle can be shorter and more features and updates can be performed – in general, it aims to bring IT development much closer to the business that is being served.

The broad goals span the entire lifecycle of a software project:

• Increased frequency of project deliveries
• Faster time to market on the original delivery
• Lower failure rate of delivered software
• Shorter lead time between business requests and fixes being applied
• Faster recovery time when failure does occur

By using a DevOps approach the IT team can maximize the predictability of new releases. Efficiency and security are both increased and the code becomes easier to maintain.

IBA Group has conducted research into the effect of using a DevOps approach to software development – although this research was focused on mainframe DevOps. These are some of the key findings:

• 20 x faster to recover when software fails
• 22% less time spent on rework to fix problems
• 30 x more deployments of new software
• 40 x lower failure rate of delivered software
• 50 x greater IT team satisfaction

All this data comes from real mainframe client projects at IBA Group. Deployment becomes more reliable and more frequent when people work together using this type of framework. The IT team uses a form of system thinking, which really means that they create a culture of shared responsibility for the project. This culture encourages transparency and shared responsibility – problems that one team member may have hidden in a regular organization are shared and managed together.

Automation is also an important part of the DevOps culture. The aim is to automate all the routine tasks a developer usually needs to manage. This also creates a far more satisfied developer who can focus on the more interesting and challenging parts of the project. This naturally leads to better quality and performance – enhancing the reputation of the team.

Most executives outside IT are not really familiar with software development practices, but it is becoming more important to understand because a different approach to the way that software development is managed leads directly to business effects, such as better quality, fewer failures, and a team with higher satisfaction in their job. DevOps is well established as a practice with a decade of conferences and articles all exploring how it can be used effectively. If you need to purchase any form of software development from an IT company then how they are managing DevOps should be one of the first things you ask.

Author: Mark Hillary

In my last couple of articles here I explored Cloud Management Platforms (CMP), both to define what they are and to give some ideas on how to choose the best one for a particular business. It was therefore interesting to see that a recent article in ITPro Today explores the top ten CMPs available today, assessing all their strengths and weaknesses.

All CMPs need to provide lifecycle management – this is the ability to track cloud resources over a period of time – and data protection, in addition to the main functionality of controlling and automating cloud-based processes. John Webster, a senior partner and analyst at the Evaluator Group, managed the research published in ITPro Today. One of his main concerns when comparing the different CMPs was that not all of them are keenly focused on data protection – they are focusing mainly on basic functionality.

John explained: “Data protection and disaster recovery is an IT responsibility, a bedrock function, and I think that the vendors in this space have to really start looking at that seriously.” He added: “Vendors will likely provide these capabilities through extensions to data protection and disaster recovery applications that are already available in the market.”

The top ten list of CMPs was created by weighing up several factors, including:

• The quality of the user interface
• Ability to manage various groups of users
• Complexity of the service

Cost control is an important area of functionality that most customers want to use, but many CMP vendors find it very difficult to offer because they are constantly updating their products. The ability to create efficiency is much easier to plan when the software is stable. If the CMP is constantly being improved then there is an almost constant need to explore efficiencies.

John explained that support for Artificial Intelligence (AI) is likely to be an important capability in the near future. He said: “Support for cloud-native including Kubernetes, and application migration will be key functions in cloud management platform tools. AI assistance, or the assistance of artificial intelligence, will become more and more important as time goes on.”

Some of the CMPs on the market at present have been built from trusted and tested systems that were essentially managing IT estates – they have been converted to managing cloud-based systems, but others are built from scratch. It’s important to be aware of this when selecting a supplier. The start-ups might move faster and add more features all the time, but their platforms may be less stable and less tested in real-life situations.

The ten CMPs are not listed with the best in position one; they are just grouped as the ten best CMPs. This is because ultimately the right choice of CMP will be based on the different priorities and needs of each company. Follow the link to the article and you can read the top ten free, providing you submit your contact details to the magazine.