top of page
  • Madeline Weiss, Director


Advanced Practices Council members – senior technology and data executives across industries – gathered virtually in February to continue learning from exemplary researchers and practitioners (including themselves) on topics they voted as high priority for their future success.

The meeting was about rethinking – ways to add value to organizations by removing traditional constraints and expanding the boundaries of the possible; by reinventing customer experiences; and by attracting needed talent.


Dr. Kristian Hammond of Northwestern University reviewed the many uses of AI, including machine learning, natural language processing, deep learning, facial recognition, voice recognition, process optimization, recommendation systems, robotics, self-driving vehicles, personal assistants, and predictive analytics. These AI driven capabilities are needed for many functions within the organization, including fraud detection, customer qualification, supply chain optimization, identifying market and customer insights, automated communication, price and sales predictions, cyber-security, resume filtering, employee tracking, and investment decision support. Hammond learned from a recent survey that organization leaders perceive AI as game-changers that can provide competitive advantage through efficiencies, increase productivity, reduce factory failures, increase customer loyalty, and allow for better data-driven decisions. However, Hammond pointed out that over 85% of AI projects never get beyond proof-of-concept because companies have insufficient relevant data, have unclear tasks targeted for AI adoption, and have inadequate coordination across operational functions. Organizations must build systems that are integrated into business goals, workflows and existing infrastructure while attending to the source and quality of the data. He suggested five proven steps to deployment beyond proof-of-concept: (1) build tools that support the task done by humans to make the data and process explicit; (2) track user actions and tie them to conditions under which they were taken; (3) use the data to train an automated system, tying conditions to actions and decisions; (4) use the tool to test the resulting intelligence against user actions; and (5) transform the tool from one that supports user action to user verification. He cited an insurance company that uses AI to screen thousands of daily email claims and routes each to the appropriate location. The system replaced 15 people. Hammond asked us to rethink the people needed to create valuable applications using AI. Instead of searching for one individual who can do it all, he suggested thinking about a team of people who together have capabilities to define the business problem, identify the needed AI architecture and tools, develop or contract the applications, and manage the development.


Dr. Sam Ransbotham of Boston College and his colleagues conducted surveys seeking cases where organizations have gained significant financial benefits in 2020 from scaling machine learning initiatives. Ransbotham reported that only 11% of respondents could report significant financial benefits from scaling pilots. Ransbotham identified the factors that led these 11% to the benefits: (1) appropriate infrastructure to support machine learning; (2) employees who understood the business as well as the technology; (3) strategy geared towards solving business problems; (4) aligning production of algorithms with the ability to understand and use their outcomes; and (5) thinking creatively about solutions. Ransbotham cited such financial institutions as Fidelity, HSBC, and Barclays for thinking creatively about solutions to customer verification. An obvious solution would be to employ chatbots to gather information from customers who call; a less obvious but more customer-centric solution would be to have humans answer phones and have AI systems in the background conducting verification by assessing phone latency and matching voice prints. They selected the less obvious solution and gained customer acceptance (human answering their call) as well as speed (20 seconds faster) and accuracy.


After Dr. Didier Bonnet of IMD reminded us of the traditional way of conceiving a hotel experience, he asked us to reimagine it. What if the hotel provided a contactless experience using digital technology to check in via biometrics, escort you to your room via robot, open your door, turn on lights, adjust the room temperature, and operate the TV? Leyeju Smart Hotels created such hotels at significantly reduced costs to the company (40-50% reduction) and hotel guests (room rates of $30-$60). It required pivoting to a digital mindset.

How can digital mindsets be created? The first step is to avoid common mindset pitfalls: (1) stick with assumptions common in the firm’s industry (e.g., customers buy insurance for a period of months or years vs. for a day’s excursion); (2) view digital technology as a substitute for previous processes (e.g., digital menu on wall as substitute for paper menu on restaurant table vs. offering nutritional data, meeting the chef, or offering wine pairings); (3) hold onto old way while introducing new digital way (e.g., sell old technology along with new vs. letting go of the old); (4) gather customer input by talking with customers (e.g., hold focus groups vs. gain contextual data at the point of use). These pitfalls can be avoided through zero-based design.


According to Kforce executives John Megally, Manish Mohan, and Igor Rakhmanov, organizations are fighting a war for technology staff. Candidates have many attractive employment choices without having to physically move and organizations must sell them. CIOs and their leaders can win by stressing the purpose or mission of their firms, presenting opportunities the candidates would have to grow their skills, assuring that the firm takes care of staff, and demonstrating how unbureaucratic the firm can be (e.g., making hiring decisions quickly). The Kforce executives suggested that CIOs cast a wide net for candidates, such as those who don’t have typical technology degrees or background. In addition, Kforce executives advised preparing in advance by anticipating needs and hiring before staff are needed (before other firms lure good candidates) and by adjusting salaries of existing staff so that new hires will not receive inordinately higher salaries due to market pressures.

38 views0 comments
bottom of page