||Conference Home||Register Now
Executive Track Keynote Presentation
A common organizational challenge
Any organization is only as strong as its weakest link and only as good as the decisions its top management and technical experts make. Harvard Business Review states that decisions within a company lies on a spectrum.1 On the one side there are the many simple, frequent, tactical planning, execution and improvement decisions made by experts that have cumulative value over time. On the opposite side you have the large, complex, infrequent strategic decisions made by top management that can make or break a company. In between these extremes, there is a large middle region where decisions are both frequent and individually significant. What makes these decisions difficult to get right and to scale to ensure the ongoing improvement and growth of the organization, is that they require subjective judgments based not only on data and reasoning, but also, based on intuition and experience.
Unfortunately, the availability of the best experts that get such subjective decisions right (at least more right than the rest), is almost always a scarce resource. As such, the best engineers, doctors, lawyers, psychologist, sales people etc. will always be overloaded. When such tough decisions are made by either the overloaded top experts, or delegated to those without the necessary intuition and experience, the speed and quality of decisions and therefore the performance of the whole organization suffers.
Bad decisions that occur in this middle spectrum also results in firefighting that waste the limited attention of top management – the true constraint in any organization. This in turns distracts them from focusing on the large strategic decisions they should be focused on.
Imagine if you could clone your top expert’s decision-making ability to review, not only a few of the critical decisions made every day in this middle spectrum, but to review every decision in this spectrum that could benefit from their intuition and experience.
Past Solutions that have only partially succeeded
Organizations have tried many different ways to improve the decision making in this middle spectrum. They tried to use their top experts to do training, to document their best practices, to create decision trees and even tried AI to find the optimum decision. Yet, in most real-life environments, the top experts outperform the rest of their peers and AI engines, by an order of magnitude.
What if we did not try to use AI to mine big data, but rather to mine the top experts to extract the Natural Intelligence they have developed over many years.
This is exactly the direction that the speaker followed. Using this approach, he created TOM – short for Tacit Object Modeler - which uses a combination of most advanced Machine learning and other Artificial Intelligence methods to learn how your top expert makes decisions.
Once their decision process has been captured in TOM, it can then be scaled and made available to support decision making by employees throughout the organization and used as field support, a virtual advisor, or even a training tool. TOM can increase your expert’s availability and decrease the demand put on top management.
In this presentation, the speaker will provide an overview of the initial triggering event that provided the "ah-ha” moment, how the TOM technology was developed and tested in the field, sharing the results of a few specific cases where TOM was used the clone the top performers from mining, insurance, airlines and banking and end with their view on where TOM can be applied in the future.
CARL WOCKE is one of the leading experts in the world on Predictive Analytics and Artificial Intelligence. Carl is also working with Dr. Alan Barnard, CEO of Goldratt Research Labs on a number of projects where Theory of Constraints is being applied in conjunction with AI and Predictive Analytics.
Carl has developed a methodology to model subject or domain expertise in a machine readable format. This methodology effectively facilitates the transmission of subjective knowledge into a logical, transparent framework. The process starts with the creation of a domain specific environment (simulator) with which the expert interacts. The interaction is iterative in nature and is designed to decompile the interrelationships between the considered variables (input variables) that are used by the domain expert in a decisioning process.
The outcome of the process is the ability for computer software to exhibit subjectivity at rule level. This effectively means that we are able to generate software that reacts in the same way as a particular individual, imparting value to a process as if the individual were an active participant.
Currently is the founder and Managing Director of Merlynn Intelligence Technologies – a company on the forefront of human to machine knowledge transmission using machine learning and artificial intelligence techniques. He is considered as one of the pioneers in building technologies that enable human based reasoning within machine environments.
||Conference Home||Register Now