Taking the Right Technology Initiatives
Only IPv6 has the Backbone to Carry the IoT
The Rising Tide of M2M and IoT Applications
M2M in Healthcare: High Market Potential, High Human Potential
Toward a Services-Oriented Business with Connected Machines
Doug Bellin, Global Lead, Manufacturing and Energy, Cisco
Machine Learning Will Power the Next Decade of Enterprise Software
By Jesus Rodriguez, Managing Partner, Tellago, Inc
Machine learning is rapidly becoming one of the most important trends in the enterprise. A combination of Moore’s law, the rise of big data, and the evolution of technology stacks have finally delivered the promise of machine learning technologies for many enterprises. However, machine learning extends beyond a standalone industry trend and has the opportunity to power the next wave of innovation in the enterprise.
The last decade has seen a renaissance of innovation in enterprise software powered by movements like the cloud, mobile, and big data. With these trends established as mainstream technologies in the enterprise, the market is turning its attention to technologies that can become the “next big thing” in enterprise software. From the lead technology trends in the market, machine learning appears to be on a trajectory to power the next wave of innovation in the enterprise.
Beyond the obvious technical value proposition, a transformational enterprise trend has to focus on a large market, enable simple distribution models, and enjoy a certain level of maturity from the prospective enterprise buyer.
“From the lead technology trends in the market, machine learning appears to be on a trajectory to power the next wave of innovation in the enterprise”
Causes of Rise of Machine Learning in the Enterprise
Many factors contribute to the adoption of machine learning technologies in enterprise environments. The rapid evolution of the machine learning stacks, adoption of big data technologies, as well as the explosion in volumes of data processed by organizations are some elements that are conspiring to embrace more advanced data analysis techniques.
Moore’s Law and the Evolution of GPUs
Machine learning algorithms have benefited a great deal from the emergence of graphic processing units (GPU). Given the compute intensive nature of many machine learning algorithms (particularly deep learning) they tend to effectively take advantage of GPUs and their performance improvements compared to traditional CPUs. In that sense, GPUs enable the execution of machine learning models that were considered impractical just a couple years ago.
Machine Learning Stacks are Finally Useful
After almost 20 years of evolution, machine learning stacks have finally reached a point where they are being widely adopted by developers and incorporated into third party applications. In that sense, some of the machine learning platforms and frameworks based on technologies like R and Python enjoy large and active developer communities, accelerating the level of innovation in the data science space.
The Emergence of Big Data Platforms
The explosion in volumes of data stored by organizations has drastically improved the viability and efficiency of machine learning models. At the same time, the emergence of big data platforms allows organizations to store and process volumes of data that were unthinkable a few years ago. These improvements in data storage and computation have had a major impact in machine learning applications. As to be expected, machine learning supervised and unsupervised algorithms tend to be more effective as they process larger data sets.
Horizontal and Vertical Evolution
Many enterprise software trends experience specific growth as a platform before the creation of industry specific applications. The big data movement is a perfect example of this pattern. The space was initially dominated by infrastructure platforms like Hadoop distributions which, after going mainstream, began powering a new generation of vertical big data solutions.
The machine learning space is deifying that pattern and experiencing significant developments in both horizontal platforms and vertical applications. At the same time we are seeing a new generation of machine learning platforms providing the infrastructure for building modern data science applications, we are also witnessing new leading SaaS applications expanding their capabilities by leveraging modern machine learning techniques. This duality is allowing the machine learning space to evolve faster than other hot enterprise software trends.
Every Major Enterprise Application Trend Will be Redefined Using Machine Learning
The transformational impact of machine learning could be greater than anything ever experienced in the enterprise space. From some perspectives, it can be argued that every major software trend in the enterprise space will be redefined using machine learning as the foundation. In fact, this is already happening today in different sectors such as the ones explained in the following sections:
Best minds of our generation are spending their days finding better ways to place advertisement. That’s just one of the best examples of how machine learning is transforming marketing. Predictive lead scoring, or intelligent ad and content placement are some of the new and popular marketing techniques that actively rely on machine learning models.
Forecasting analysis, customer sentiment analysis, and customer churn predictions are some examples of machine learning disrupting traditional sales processes. These techniques are starting to be included in traditional sales tools such as CRMs and ERPs to create more intelligent sales processes.
Machine learning is powering the next generation of innovation in the enterprise security space. Techniques like security threat analysis and malicious pattern recognition are actively used in modern security platforms to provide more insight concerning potential security risks in enterprise operations.
Financial technology is being completely disrupted by the emergence of data science and machine learning. Equity investment, high frequency trading, financial planning, etc., are some of the most innovative use cases that are leveraging machine learning in the financial industry.
The lead platforms in the application performance and operational monitoring space are starting to leverage machine learning to obtain additional insights about system logs or application activities. Additionally, many of these platforms are starting to leverage machine learning to proactively predict failures connected with specific business processes, and adapt accordingly.
Building on the Shoulders of Giants: Powering the Next Decade of Enterprise Software
Despite the rapid growth of enterprise machine learning solutions, they are still only grasping a very small portion of the overall market opportunity. The next decade should bring us a proliferation of machine learning solutions ranging from infrastructure platforms to new lines of business capabilities. As an enterprise software trend, machine learning appears to enjoy the unique characteristic of being accelerated for equally fast growing trends like mobile, cloud computing, IOT, and big data. Each one of these trends is becoming, in their own right, foundational building blocks of enterprise software solutions while, at the same time, surfacing new scenarios and simplifying the adoption of machine learning solutions.
From a market development perspective, machine learning is evolving into what can be considered an ideal model. The market is not only seeing the evolution of machine learning solutions as both infrastructure platforms as well as industry-specific solutions, but both trends are also being accelerated by the adoption of relevant technology trends like cloud, mobile, big data, and IOT. If this pattern continues, machine learning will become as ubiquitous as databases are today. In any case, machine learning is advantageously positioned to power the next decade of innovation in enterprise software.