Executive Interview: A Look Back (and Forward) at 20 Years of Decision Management
cheryl wilson 270003VHSH email@example.com | | Tags:  decision-management business-rule-management brms business-rules
0 Comments | 4,317 Visits
In light of the buzz around the practice of decision management and the release of a new IBM product in this space – coming October 6 – I had the opportunity to interview the former CEO of ILOG, Pierre Haren. ILOG was acquired two years ago by IBM. Pierre Haren was the founder and CEO of ILOG for 21 years, introducing ILOG on the NASDAQ 10 years after its start and driving it to $200M in revenue by 2008. When it was acquired by IBM, ILOG was the world’s leading company for Business Rules and Operations Research tools.
CW: For the benefit of our visitors who may be not as familiar with the emerging practice of decision management, how do you see it?
PH: Decision Management for me is the activity of transforming decision processes which happen in human brains into actionable objects that can be edited, traced and executed so that part or all of the decision process can be automated.
In a sense, BPM is part of Decision Management because Decision Management is about breaking down a complex process such as "attributing a loan" into a set of smaller, simpler processes based on the key decisions that are driving them, such as "deciding if the customer is eligible for a loan" then "deciding the duration and amount of loan that the customer is entitled to" then "deciding the best credit rate we can offer profitably to that customer" and so on.
Breaking down a complex process into smaller parts helps the understanding, the possible automation, and the traceability of the business decisions. Of course, business rules have been used for a while to describe and automate smaller fragments of decisions like those inside the "deciding the best credit rate we can offer profitably to that customer" process, for example. And many other technologies can be brought to bear to improve decision making such as segmentation with predictive analytics, optimization with operations research tools, and so on.
CW: From your perspective, how did we get from the first rules engine to an emerging practice that brings together several technologies like business rules, events and analytics?
PH: Many of the pieces of decision management have existed separately for a long time. Business Process Re-engineering is over 30 years old, so are work-flow engines, rules engines, data segmentation through statistics. Operations Research is over 60 years old. What is new is the widespread understanding that all of these pieces can complement each other, as well as the technical capability of SOA and, more recently, Cloud, to bring all these technologies to bear on the same problem for a cost that is much lower than the benefits that can be derived.
CW: Why do you think decision management is important for businesses today? Why should line-of-business care?
PH: It is all linked to the acceleration of the rate of change in the world.
In a very stable business environment, business users can specify a decision support system and have their IT department code it in Cobol or Java within a year. Human operators outside the system can absorb the minor changes that are inevitable. Alternatively, business users can also import best-of-breed predefined processes by deploying ERPs, for example.
However, when business models change at the current rate, within a year, a month, sometimes a day, it becomes impossible to rely on such stable systems. So, line-of-business people who are the first to know which changes have to be made are looking for solutions to modify their "decision management" support systems at the rate imposed by the world.
CW: Can you speak to any defining moments over the past 20+ years of being in this business? For example, any customer implementations that stood out as shining examples of the power of decision management?
PH: We had a great moment when we were able to successfully deploy a complex mix of ILOG and IBM tools at IBM Fishkill, in what was considered the most modern and fastest changing semi-conductor plant in the world. At that time, before our acquisition by IBM, IBM was a reluctant customer. I visited the head of the plant and promised that we would install the complete system for free. He would only pay if he kept the system in operation for more than a month.
Understandably, he was skeptical that a small decision management company could improve on the work of hundreds of excellent engineers who continuously tuned the $4B plant. We deployed a system which every five minutes re-computes the optimal plan for the whole plant for the next eight hours. It combines real-time data from the MES system to rules-based processes, an optimization-based scheduler, and advanced graphics to enable the plant operators to "see the current future" for the next eight hours.
The first day's objective was low: do no harm. Within a month, all the operators where convinced and we were paid. More important, we changed forever the balance of what computers do and what operators do in that plant.
CW: Where do you think the practice of decision management is headed? In the next 5-10 years?
PH: We are sharing more and more of our daily workload with computers. It’s not a fixed boundary, computers can do more things at a time when the world is more complex due to regulatory changes, catastrophes, and the rapid evolution of technology and competition.
I still like the insight shared by the technical director of Fishkill IBM plant when we asked what the operators were doing since our automation scheduling system was deployed: "We’re no longer running the plant, we’re observing the plant, optimizing it and solving any issues that arise.”
So, decision management enables a migration from doing the work to observing computers doing the work, optimizing what they do, and adjusting it to the issues that arise. This trend will continue for the next 10 years. We’ll then apply decision management to decision management itself, and automate the supervision role, segment the issues and the appropriate responses, and so on. This is a long journey.
CW: What's the best advice you ever received from a customer over the years that would be helpful to any companies considering the adoption of these technologies?
PH: Alain Morel was the CIO of SNCF Freight, which is the arm of the French national railroad dedicated to the transportation of goods rather than people. They were facing a lot of issues with their billing system, and decided to automate the link between contracts and billing to reduce the number of errors and customer dissatisfaction. A team started to implement a rules system, then they stalled the implementation for six months.
We were eager to help and Alain explained: "After a few weeks of the implementation of the rules application, we discovered that sales, marketing and accounting don’t share the same vision of what is a contract at SNCF-Fret. So we’ve stopped the rules implementation process and we’re in a discovery process to agree on what we’re doing.”
To me, this summarizes the advantage and the difficulty of deploying decision management systems in any organization: the implementation process surfaces many hidden contradictions or battles between people or departments, which trigger errors and customer dissatisfaction down the line. These issues can go on forever as long as they’re happening in people's heads. Describing them and then automating them gives a strong scrub to the existing situation, and leads in all cases to very positive results for the business.
CW: Any last words of practical wisdom?
PH: Make sure you implement these systems before your best experts retire because the discovery and testing of decision management systems require excellent knowledge of the company, its past evolution and projected future. And beginners typically don’t have this in their heads.
To learn more about Decision Management, click here.