Machine learning models can surpass human capability in coping with significant volumes of data, finding high-order interactions and patterns within the data and dealing with highly complex business problems. Digital businesses will adopt machine learning in more and more use cases, driven by the availability of sensor data, expanding bandwidth and sinking storage costs.
Many people believe that machine learning is all about automation and eliminating human input. However, successful machine learning relies on talented data scientists able to orchestrate a whole range of computing infrastructure to extract knowledge from data.
While the basic concepts of machine learning have been around for decades, interest is at an all-time high. Digital businesses are increasingly adopting machine learning, driven by the availability of sensor data, expanding bandwidth and sinking storage costs. There’s an abundance of examples across organizations:
- Sales and marketing– Machine learning models used for product recommendations are constructed to predict which product a customer is most likely to buy. They take a customer profile as input (customer activities, recent purchases and personal information) and map this to the predicted likelihood of the customer responding to a given offering.
- Risk and fraud management– Machine learning in fraud detection is typically used to map descriptions of transactions to their likelihood, indicating whether an ongoing transaction has a high likelihood of being fraudulent. In credit risk, it can map the loan applicant’s details (demographics and credit/payment history) to the likelihood of them defaulting on the given loan.
- Smart transportation– Traffic optimization can be achieved through an understanding of traffic patterns using sensor data, accidents and roadworks. A machine learning model predicts delays or road obstructions and recommends a faster route for public buses, as well as consumer and commercial vehicles.
- Supply chain processes– Machine learning models in asset performance management take the operating conditions of assets, such as wind turbines, solar panels and nuclear reactors, as input and predict when failures will occur. The objective is to decrease maintenance costs and minimize downtime.
- Healthcare– Machine learning models in early-warning systems for employees will analyze sensor data in hazardous environments — such as measurement of air quality, equipment performance or employee productivity, or even atypical behavior — to predict the likelihood of accidents. This application has been widely adopted to alert truck drivers to potential accidents.
The increasing complexity of problems has, at best, increased the time needed for system architects and solution engineers to fully understand a problem. At worst, it makes it impossible for them to solve the problem. Machine learning provides a methodology for dealing with the complexity of digital business scenarios in a number of ways:
- It copes well with data variety because of its native data fusion capabilities — where many different data sources can (“simply”) be concatenated into an input vector.
- It uses predictive cues in the data and is very useful for problems where the data would be too noisy and highly dimensional for a human expert or traditional software to see any meaningful pattern.
- It doesn’t require someone to fully understand the problem conceptually. In the example of automatic car steering, the physics of steering is known but the relationship between the front-view of the car and how to steer the wheel is far from trivial.
- It offers more consistent solutions than human solutions in situations where consistency is critical, such as customer engagement, credit underwriting or quality monitoring.
- When placed against human learning, machine learning has the ability to cope with much larger datasets, and to update its knowledge on an ongoing basis by taking into account new factors.
Accelerating Digital Transformation using AI
The question being asked by chief executives around the world is not if digital disruption will occur, but what it means for their business. Perhaps more importantly, organisations are considering how they can leverage digital transformation to advance their competitive position and improve performance. Increasingly, it has become apparent that artificial intelligence (AI) may very well hold part of the answer to those questions.
There are three primary ways that AI is being used by market leaders to accelerate customer-centric experience design and achieve digital transformation – insight generation, customer engagement and business acceleration.
Insight generation involves extracting meaningful and actionable intelligence from ever-increasing quantities of available raw data. With the amount of information in the world nearly doubling each year, it is no surprise that data complexity is the top challenge standing in the way of digital transformation, according to preliminary results from a study by FORTUNE Knowledge Group and Publicis.Sapient.
AI tools are able to perform what no single human or even team of people could hope to do; they can read, review and analyze vast quantities of disparate data, providing insight into how customers feel about a company’s products or services and why they feel the way they do. Luminoso, an AI company with its roots in MIT’s Media Lab, has built a robust business performing precisely this task.
Customer engagement has long been the Holy Grail for marketing and customer relationship management programmes. Today, AI is radically enhancing the personalisation of information that fuels such engagement. Nowhere is this more evident than in AI’s next big thing – chatbots and virtual assistants.
Multiple companies, such as Viv, Facebook and Nuance, are providing frameworks and turn-key solutions in this space, allowing for services as diverse as media content distribution to customer service support and customised marketing campaigns. While the technology advances are exciting and bode well for business application, successful use cases will be grounded in a strong user-centred design process, leveraging the input of business and marketing experts as well as those of the IT division.
Business acceleration refers to how companies use AI to expedite knowledge-based activities to improve efficiency and performance. Examples range from hospitals finding potential patients for drug trials to financial institutions creating investment strategies for their investors.
While these types of activities are often viewed as opportunities to reduce costs through the automation of internal processes, they also should be considered in terms of their ability to transform the customer experience.
For example, if a bank can use AI to reduce the time it takes to approve a loan, it not only reduces its own costs, but also provides an improved customer experience. As a result, when AI tools such as Watson from IBM and Cyc from Cycorp are deployed, market leaders ensure they leverage the technologies with both cost-cutting and customer satisfaction in mind.
Beyond digital transformation: from holistic optimization to the optimization of the living ecosystem
Technology is never the final answer nor the essence. It’s what it can do which continues to matter. All the rest is human. And that’s also what intelligent information activation is and should be about: smarter, better, more valuable and even life-saving outcomes for people.
Indeed, holistic and hyper-connected optimization, as far as we’re concerned the goal of information but also of technology, disruption and digital transformation as such. Holistic optimization with less focus on the technology dimensions as it exists now and more on the human/emotional aspects are what’s coming AFTER digital transformation.