Point of View

The future of decision-making: augmented intelligence

Today’s AI doesn’t just shuttle passengers and cargo. It wins races.

  • Facebook
  • Twitter
  • Linkedin
  • Email
Explore

The future of decision-making involves a creative mix of data, analytics, and artificial intelligence (AI), with just the right dash of human judgment. The result is augmented intelligence – where the analytical power and speed of AI takes over the majority of data processing, guiding human employees to make more agile, smarter decisions and find new discoveries.

This is in stark contrast to just a decade ago, when analytics was considered a mere back-office advisory function. Now, it has reached a level of maturity where it is happening all the time in our everyday software and processes. Think of Apple’s Siri. You might not realize a lot of data and analytics go into producing its responses and suggestions.

The maturity of analytics has captured the attention of the heads of major companies. Yet in spite of advancements over the years, few have been able to keep up with how to use analytics and AI among employees, in processes, and with proper oversight. The result is a lot of good ideas and technologies, but applications that fall short of their potential.

If you think about analytics in terms of cooking, then data and technology investments, like machine learning, are all amazing ingredients, but you need fine chefs to make a Michelin-worthy dish. These are your data scientists, data engineers, and people with domain knowledge that can serve as translators, or “bilinguals,” and talk domain and analytics at the same time.

Equally important, you need to have the right models and processes in place, i.e. the recipes for success. Today, you can plug data into AI, and it can create its own models and make predictive recommendations. But these models do not exist in a vacuum. They involve inputs and outputs that impact the rest of your business. You have to think about how these models fit in and how to prioritize the insights from data. Moreover, you need governance over augmented intelligence to see that the automation is working and people know their role in the new man-meets-machine workforce.

If you are investing in analytics and AI, then think beyond data and technology. You need bilingual talent to bridge the gaps between industry and technology, and develop the right solutions. You also need to consider the paradigm shift in how man and machine will work together. Only then can you serve up a winning application and get the
most value out of investments.

Industrializing analytics: why bilinguals are important

There have been significant advancements in business analytics over the years, yet applications are still considerably less practical. Efforts to harness the ever-growing volume and variety of data into automated processes  – the “industrialization” of analytics – is a mixed blessing that is leading to a widening imbalance among talent.

There is a deep bench of data scientists with the ability to spin extensive webs of data into theoretical models. But there are far fewer people who can take a model and translate it into a working reality in business and on an industrial-size scale.

A decade ago, in a period we’ll call Analytics 1.0, there was a talent group with one set of skills and tools in the back office, i.e. statisticians. They would often use statistical sampling because much of the necessary data wasn’t available, come up with probability distributions, and develop a model. The model would be handed off to the IT team, who would take the model and development specifications for deployment. For example, if someone was applying for a credit card or a mortgage, the IT team would use the developed model to create something to approve or deny applications.

Today, in Analytics 2.0, many things have changed and evolved. The sheer amount of data has grown. The process has evolved far beyond statistical models with the advent of AI that can sort through data patterns. Most of all, the profile of the people working with data and technology has changed. You are much more likely to find two separate camps: industry experts with domain knowledge built up from years of experience or technology experts who may be well-versed in data, analytics, or AI. The ability to understand both domain and technology is difficult to cultivate, which makes a bilingual talent, who is capable of operating seamlessly with equal proficiency in both camps, highly desirable.

Take a copy for yourself

Download PDF

If a project has 15 domain experts but no data engineers or bilingual team members, the odds of a successful analytics model deployment are not good. Case in point: A bank in Charlotte worked for two years to deploy a single production model. However, the team of banking specialists who  developed the model lacked the practical data engineering skill set to scale it up, thus other prospective deployments stalled.

At Genpact, our bilingual talent combine deep domain knowledge with technical expertise – from data science to big data engineering and AI – to embed advanced analytics into our clients’ core business processes.

Augmented intelligence: when man meets machine

Most enterprises have mounds of data but few insights to promote positive business outcomes. They struggle to achieve ROI, such as revenue growth, better customer experiences, and regulatory compliance, as well as to build an analytics capability for the future.

Augmented intelligence unites the strengths of people and machines when prospecting value from data. Namely, you can augment human instinct with smart algorithms that provide fast, data-driven predictive insights. These insights can help people redesign functions, detect patterns, find strategic opportunities, and turn data into action.

Intended to extend human cognitive abilities, augmented intelligence is different from straight automation. If you think about it, most processes in the future will be designed for straight-through processing, where there will be no humans involved in the process. Currently, that is not possible because in 25-30% of cases, you need humans to step in.

Think about an aircraft autopilot. In modern aviation, the autopilot can operate independently, controlling heading and altitude, or it can be coupled with a navigation system and fly pre-programed – once the aircraft has successfully become airborne. You still need a pilot for takeoff and landing – for now. In a perfect world, an autopilot system would incorporate human knowledge, plus experience  focused through the prism of intuition. The result would be human judgment extended via augmented intelligence.

To fully appreciate the benefits and potential of augmented intelligence in analytics, it is critical to  completely reengineer your mind set. You are not designing a world that is predominantly manual with 30-40% automation. The goal is to create an entirely new process – in a world that is predominantly automated and designed for 20% manual exceptions.

Governing analytics: looping feedback

In today’s businesses, almost all employees are now knowledge workers. Yet many companies continue to struggle with providing the right set of tools to support their knowledge workers. In this era of information overload, these employees increasingly need and want to be able to make faster and more complex decisions using large amounts of available data.

Augmented intelligence can be considered such a tool. The application of augmented intelligence isn’t about replacing humans but rather to find hidden meaning within data by uncovering patterns and correlations at a massive scale. Businesses will have access to fertile new ata. New applications will emerge. And decision-making and actions will improve, provided you have feedback loops built in for continuous improvement. Feedback loops are important for improving upon algorithms and in making sure that when things do not happen as expected, there are mechanisms in place to understand why.

The future of decision-making presents a transition into a whole new world and way of thinking. Melding data, analytics, and AI for augmented intelligence is a critical and necessary – place to start. But, ultimately, technology, as powerful as it is, is just there to help. Success requires the right talent, mix of human instinct with machine algorithms, and oversight over technology. It’s time to start cooking and taking flight – or risk being left behind.

Visit our analytics page

Learn more

To effectively attain augmented intelligence, consider this checklist of five key areas:

  • Cultivate and look for the right talent, i.e. bilinguals with domain and technology expertise
  • Have proper governance over automation and people
  • Think through change management, specifically ensuring a smooth adoption in the new ways of work and upskilling talent
  • Make sure your technology and processes have continuous human input
  • Develop white-box algorithms that can explain why things are not working as they should