3 Uncommon Examples of Deep Learning in the Enterprise


By Lionel Gibbons | September 11, 2017 | deep learning



uncommon examples of deep learning in enterprise.jpgWhile tech companies like Google and Facebook have already invested heavily in deep learning technology as a core part of automating their services, the broader impact on enterprises more generally is just beginning. According to Gartner’s “Predict 2017: Artificial Intelligence,” startups will overtake Amazon, Google, and other leaders in driving AI disruption in the enterprise.

While deep learning in the enterprise is still at the leading edge of its adoption beyond research, enterprises and startups are now harnessing deep learning and AI in ways that hold the potential to change entire sectors.

The innovative ways that deep learning is transforming enterprises are uncommon at this early stage, but here are three ways that are gaining ground.

#1: New Uses for Intelligent Image Analysis in Healthcare

In a recent Stanford University Research report by artificial intelligence scientist Sebastian Thrun and colleagues, a “deep learning” algorithm is now capable of diagnosing potentially cancerous skin lesions as accurately as a board-certified dermatologist. This is a leap forward for the more common image recognition capabilities of deep learning, and points to a burgeoning field of potential healthcare uses going forward. Radiology as well as early-stage cancer and diabetes complication detection are considered just some of the major drivers of deep learning within the healthcare enterprise moving forward.

#2: Deep Learning in Marketing Advertising and Sales in the Enterprise

Every enterprise working in the B2C space must find ways to more accurately predict buying habits and influence purchases through targeted marketing to rise above the noise. While typical personalization models were once enough, enterprises are now turning to new deep learning algorithms to detect unexpected situations and hidden potential.

This essentially comes down to more reliable, richer, machine-interpretable user descriptions of customers’ buying potential without the need for human expertise. Deep learning algorithms are making it possible for enterprises to define every potential client product search and identify behavior changes dynamically. This results in precise conversion probabilities for every online search by every user.

While fortunes are already being made by leaders such as Amazon using deep learning algorithms such as its anticipatory shipping, deep learning is the perfect tool to predict a user’s desires in the advertising industry.

As part of marketing and voice of the customer (VOC), deep learning is being combined with machine learning, Internet of Things (IoT), blockchain, analytics, Big Data, and data intelligence into a holistic digital innovation system in the B2B space as well. SAP is now offering enterprise sales divisions its Leonardo Machine Learning software, which enables self-running customer service, discovers churn indicators as part of customer retention, and recognizes brand imaging for accurate marketing spend tracking/adherence.

#3: Enterprise Application Development

Application development is being revolutionized by deep learning as part of the Low-Code/No-Code space by enabling wizards that can discern the intent of the application creator, thus fleshing out as much of the completed application as they can automatically. This holds the potential to reduce a typical six-month application design lifecycle requiring dozens of people and millions of dollars. Deep learning can be used to cut that down to just two weeks, three people, and less than one hundred thousand dollars while delivering a faster, higher-quality, and more flexible app than could be conventionally produced.

This will be a slower transformation because of pushback from system integrators, IT departments, and even DevOps teams that see it as encroaching on their freedom and ability to make money, but the reality is that it is likely an unstoppable disruption. Of course, there are other challenges as well as opportunities posed by deep learning and application development in the enterprise. A recent article on DevOps.com discussed the sea change for developers surrounding where data resides and the use of the cloud as part of DevOps.

As the race heats up for enterprises to harness deep learning through investment in platforms and systems that can harness big data, they are being fueled by companies such as the inventor of the GPU, NVIDIA. Many of the leading providers are partnering with NVIDIA to deliver highly agile solutions to tapping into big data, which is crucial.

For example, leading providers are offering deep learning solutions that enable infinite flexibility. That can include a one-stop shop for everything needed to spin up a deep learning environment while having the choice of frameworks, libraries, and infrastructure elements as well as monitoring and management tools in one platform. These solution providers will be instrumental in helping to propel enterprises into a previously unachievable enterprise reality in which deep learning transforms business sectors and bottom lines in innovative ways.

Building a Deep Learning Environment for Your Organization