Skip to main content Skip to Footer

Rethinking reengineering

Artificial intelligence tools can be unconventional and unpredictable, but they can also enable powerful new business processes.

B​y H. James Wilson, Allan E. Alter and Narendra Mulani


Barely a generation ago, reengineering was considered a radical (and not always beneficial) way to change business processes. Today, with the help of artificial intelligence (AI), companies can improve and change business processes automatically—an approach we call machine-reengineering.

Many organizations are machine-reengineering processes by using AI tools like machine learning—an advance in big data analytics, where algorithms better their performance by testing their own predictions on data and constantly learning from data generated by the predictions and insights.

The result? Fully two-thirds of the companies that have adopted machine-reengineering are speeding up or improving at least one process by more than 500 percent. Product development, sales, marketing, service delivery, and information and risk management are just a few of the processes that are benefiting.

Beyond extraordinary

How significant are the process changes brought on by machine-reengineering? Consider statisticians at LexisNexis Risk Solutions, who used to spend days developing fraud analysis models. Now they can achieve the same results in a fraction of the time. “We can accomplish overnight what would have taken us days to execute without advanced analytic approaches,” says Jeffrey Feinstein, vice president, analytic strategy. Manufacturers are also using systems from platforms like Sight Machine to identify flaws in their models, then determine why they are flawed and how to fix them.1 Other companies see the value of machine learning in predictive maintenance solutions.2

Businesses are also machine-reengineering by using machine learning to leverage big data. One company reduced the time analysts needed to generate a report involving up to 50 million rows of data from 40 hours to five minutes—nearly 500 times faster. The solution combined distributed computing and scale-out architecture with machine-learning technology from Paxata applied against data-science processes.

Machine-reengineering has already enabled organizations to obtain previously unattainable value and crack business problems that have gone begging for solutions. In fact, 89 percent of our survey respondents agree that machine-reengineered processes have the power to transform business models, allowing companies to be “more flexible, adaptive, productive, and to consider things we may not have otherwise,” as one executive put it.

Powerful, unconventional and unpredictable

The downside? Machine-reengineered processes present challenges on top of the project and change management concerns that executives usually face when implementing new systems. That’s because these processes optimize, repair and adapt themselves to changing business conditions without human intervention. “They manage themselves,” notes one manager. But these new processes can often be unpredictable. Nearly 80 percent of respondents say machine-learning algorithms replace planned sequences of steps with unanticipated actions. This behavior creates three new challenges for organizations.

Trusting machine-reengineered processes

Executives face a trust paradox. On one hand, employees might rely too much on algorithms for decision making. When respondents were asked to pick the top two human risks associated with machine learning, the most frequently chosen answer, at 39 percent, was that people would simply accept poor algorithms and machine mistakes.

On the other hand, said respondents, employees may be uncomfortable when machines appear to show initiative, and as a result may subvert these systems. Twenty-nine percent of respondents said they feared that people would undermine good machine algorithms and decisions. For instance, a warehouse manager who sees lower demand for a certain SKU this week might ignore a machine recommendation to order an additional 22 percent of that SKU for the following week, not understanding that the machine is optimizing inventory to meet higher predicted demand.

Identifying critical talents

Today, companies compete for analysts, engineers and data scientists who have hard technical skills and knowledge of distributed computing systems and analytical tools.3 But in the world of machine-reengineering, workers will also need other skills that could be as unconventional as the processes they support.4 These include the ability to program “belief spaces”—advanced probabilistic models that help robots deal with uncertainty—and to work well with intelligent machines.5 Mercedes-Benz, for example, has replaced large, inflexible factory robots with people and smaller, more flexible robots. Now Mercedes-Benz needs a workforce that can teach robots how to collaborate closely with employees on the factory floor.

Targeting machine-reengineering opportunities

The conventional response to disruptive new technologies—experiment on low-hanging fruit—is insufficient when it comes to machine-reengineering. Organizations must move quickly to identify and exploit business benefits so they aren’t on the losing end of a technological arms and mastery race that’s already begun. But companies must first decide which processes are the best targets for machine-reengineering and which approaches to apply.


Author Jim Wilson explains how to prepare employees for AI-powered machine-reengineering.

Dealing with the three T’s

Trust, talent and targets—executives must address all three challenges to exploit the full potential of machine-reengineering. To alleviate these challenges and gain the most benefit from machine-reengineering, our preliminary research in this area suggests a series of steps, tailored to the specific needs of individual organizations.

Overcome the trust paradox with experiments

More than ever, executives should create an environment that enables data-driven change. In such an analytical culture, employees should be expected to challenge, increase or decrease the influence of algorithms. CEO Avi Steinlauf cautions: “You can’t just turn it all over to a machine that will do your thinking for you. Tools can help, but you need smart people thinking about things in curious and intuitive ways.”6

Half of our respondents’ organizations now use prescriptive or predictive analytics, driven by machine-learning algorithms. Some of these algorithms are undoubtedly more effective than others. Underperforming algorithms should be taken out of production, while high performers should be shared and adapted across silos. In our research, a third of respondents use machine learning in only one production system. Companies should consider applications for successful algorithms in other production systems.

Scout for newly necessary skills

Many leading learning researchers are pioneering the machine-reengineering skills that organizations will need to master tomorrow. For example, MIT’s Cynthia Rudin is exploring how to build simpler and easier-to-interpret machine-learning models—insights that employees will ultimately need to make algorithms easier to verify.7

Partnering with researchers through one-to-one arrangements or via associations and consortia will help executives develop a greater understanding of emerging technical skill needs. (Machine-learning vendors can also spot emerging machine-collaboration skill needs.) Any problems that arise in a workforce’s interactions with algorithmic processes also indicate new skills to develop.

Take human aversion to algorithms: Studies have shown that people are far less forgiving of forecasting algorithms that err than of erring humans. But researchers at the University of Pennsylvania have found that if people are permitted to modify algorithms even a little bit, they are likely to continue to use them. This finding points to a newly needed management skill: Sniffing out how, and how much, staff can modify algorithms without degrading their performance.8

Prioritize projects, optimize tactics

Our respondents identified several factors executives should consider when prioritizing machine-reengineering projects, including the potential financial impact and the speed of the process change. Also important? The learning delta: the difference between an algorithm’s and a human’s capacity to learn in the context of big data. What can a machine learn that can then help people improve performance?

For decision makers, there’s a trade-off—assessing the value created by machine-reengineering processes against the acceptable investment costs and areas where customer needs and competitive conditions require faster process change. Ultimately, executives must consider their goal: Is it process optimization or business model change? Or is it to test a process that would truly disrupt the industry? Our research has identified companies that have successfully used machine-reengineering to pursue all these goals—some by applying algorithms to data that previously collected data, others by creating new data sets and APIs (see figure below).


Machine-reengineering solves problems faster by identifying patterns people fail to see on their own. With this approach, executives can create powerful new digital business processes by using machine learning and other kinds of artificial intelligence. Already, companies in many industries are seeing unprecedented results, improving processes by more than 1,000 percent and developing new business models. Executives who build an experimental, data-driven culture; recognize new human-machine interaction skills; and understand machine-reengineering’s potential beyond process optimization are best equipped to harness the power of this unconventional technology opportunity.

**Acknowledgment: The authors with to thank Accenture Institute for High Performance researchers Dr. Prashant Shukla and David Lavieri, both of whom made significant contributions to the research discussed in this article.**

1. Interviews with LexisNexis Risk Solutions and Sight Machine

2. Tomas Kellner, “Deep Machine Learning: GE And BP Will Connect Thousands Of Subsea Oil Wells To The Industrial Internet,” GE Reports, July 8, 2015


4. Industrie 4.0 at Mercedes-Benz: The Next Step in the Industrial Revolution, Oct. 19, 2015; Elisabeth Behrmann and Christoph Rauwald “Mercedes Boots Robots From the Production Line,”, February 25, 2016,

5. Ken Goldberg, “Robots with their heads in the clouds,”

6. Dale D. Buss, “The Rise of Machine Learning,”, March 24, 2016

7. Interview with Prof. Cynthia Rudin, December 10, 2015

8. Dietvorst, Berkeley J. and Simmons, Joseph P. and Massey, Cade, Overcoming Algorithm Aversion: People Will Use Imperfect Algorithms If They Can (Even Slightly) Modify Them (April 5, 2016). Available at SSRN:

Outlook: Accenture's Journal of High-Performance Business

As a showcase for the most innovative thinking on high-performance business, Outlook focuses on six core themes: Redefining CompetitivenessDigital DisruptionGlobal Operating ModelOpen Innovation, Sustainability and Workforce of the Future. We feature original content devoted to these topics as well as a selection of unique insights offered by professionals throughout Accenture.​​​




H. James Wilson

H. James Wilson

Wilson is managing director of Information Technology and Business Research at the Accenture Institute for High Performance.

Allan E. Alter

Allan E. Alter

Alter is a senior research fellow at the Accenture Institute for High Performance.

Narendra Mulani

Narendra Mulani

Mulani is Chief Analytics Officer for Accenture Analytics​.

Discover insights on-the-go with the Accenture Outlook app: