Happy Birthday, Business Rules!
cheryl wilson 270003VHSH firstname.lastname@example.org | | Tags:  brms business-rule-management operational-decision-mana... decision-management business-rules
0 Comments | 6,985 Visits
To celebrate the 25th birthday of our business rules management technology, I reached out to two IBM executives for their unique perspectives on business rules – where it’s been, where it's going and why it remains such an important business technology:
Q: Pierre, as someone who's been on the ground floor, how would you sum up 25 years of business rules technology for someone just coming to it?
Pierre: Here’s the one-minute history: Rules technology started in artificial intelligence (AI) in the 70s, then matured as expert system shells in the 80s, almost died in the 90s, and re-emerged as business rules in the third millennium. Expert Systems failed because they tried to automate experts, but there were few experts and those few didn’t want to be automated. Business Rules succeeded because it focused on the automation of simpler tasks, and there are an infinite number of simple tasks that we can automate. Moreover, expert systems required specialized knowledge engineers to program them while business rules could be edited by business users.
Most, if not all, companies want to better serve their customers with no latency and as much automation as possible, so rules technology matured by going mass market and much simpler. ILOG played a key role in this migration, carrying the technology through the winter of AI and pioneering the concept of Business Rules Management System.
Q: What's the significance of this milestone? What does it say about the technology, market -- the business of business rules?
Phil: I think it’s important to reflect on the value that has been unleashed because of ILOG business rule management over the past 25 years. Everyone associated with this technology, especially those who were with ILOG for the years before the IBM acquisition, have a lot to be proud of. ILOG led the business rules revolution, a revolution that generated a lot of economic activity and returns and helped advance the world.
Interestingly, the potential for business rules has never been greater than it is now. Pierre and his team's founding vision for business rules is in some ways only recently attainable. The role that business rules plays increases exponentially in a world of "big data," in a world of billions of connected devices sending trillions of events and messages. We see this here at IBM, where our Operational Decision Management product -- essentially the next generation of ILOG’s market-leading business rule management system (BRMS) -- brings together rules and events. We also see the value across software group from Watson to Sterling to Smarter Cities. Everyone wants (and needs) ILOG business rules.
Pierre: I would add to Phil's comments that it’s rare for the same technology to last so long and increase in relevance over time. I think this is true of business rules because a large part of human thinking happens in terms of rules.
Not to go too into the weeds, but roughly ten percent of our knowledge is about categorization and language (e.g., this is a car; this is a chair; this sub-type of chair is a stool, which can have three or four legs; etc.). About ten percent of our knowledge is about process (e.g., identify the customer > then if he is registered > then what he wants, such as Open Account, etc.). Thirty percent or more of our knowledge is about rules (e.g., if the customer account is negative, and there is another account with enough money, then do this or else do that). Most of the rest of our knowledge is about specific contexts or situations; whereby we look for previous examples which are similar to the current problem.
Categorization and Process can be harder to express, but are powerful and reusable pieces of knowledge. Rules are easier to find, more numerous, and are applied in very specific contexts so over time you tend to build a lot of them into every human activity.
Q. What have been the biggest changes you've seen?
Pierre: First, the expert system shells of the 80s had many features that have been lost in current business rules technology, such as multiple worlds’ management. For the uninitiated, multiple worlds’ management was a way to handle multiple versions of an object graph simultaneously so that you could compare different solutions automatically. It was also used for simulators to compare the outcomes of different hypotheses. For example, if you were running a road repair project, you could select the most robust solution to hypothetical “invisible road damage.” Most of these features were powerful and complex to use. In a way, the technology had to be simplified "outside" to succeed. Having said this, the performance of the modern business rule management systems is many orders of magnitude better than the original ones. Incredible, invisible algorithmic progress has been made since the initial rule engines.
I think the other big change is that since 2000 a lot of energy has been devoted to developing natural language and building structured editors for the business rules that support business users -- supporting their change management process so that many business users can cooperate to edit, refine, test and deploy a set of business rules as a Web service for example. We are now entering the world of social cooperation of business rules edition, with strong formal validation, very rapid deployment, and high frequency of change of business-rules-powered, real-time processes.
Q. Where do you think the technology is going in the next five to ten years?
Phil: I would divide the five to ten year picture into three situational categories:
1. The pervasiveness of rules. As I said before, everyone wants and needs business rules embedded as natively as possible in their applications. Because rules are everywhere, we’re working across IBM to enable this, without jeopardizing the value proposition of an enterprise BRMS. We want to address rules in a common way across the portfolio, not for internal IBM efficiencies but because our customers want and need common ways to express rules. So we’re developing a special version of our operational rules capability that will be easily embeddable in anybody's product -- inside or outside IBM. Today, IBM powers something like 70% to 80% of the world's transactions. Wouldn't it be great if in ten years we could make the same claim about operational rules processing.
2. The application of rules. Until now, a BRMS provided three primary value propositions: 1) a flexible, business-friendly "language" … 2) that can be easily and quickly changed … 3) powered by a high-throughput transactional engine. I don't mean to simplify things because there are many other value propositions around rules, but that sums up what the market thinks.
Now, we want to make the enterprise BRMS value proposition even bigger than that. First, the industry has stagnated around that second bullet: ease of use by the business users we intend to serve. If we incorporate the recent breakthroughs in usability, social and process capabilities into the product, we can really deliver quantum leaps in business user empowerment and engagement. And second, we need to actually assist in the real-world change management for this business-led change. Up until now, the industry has thought that once a business user changes a rule, the hard part is over. But business rule customers tell us just the opposite: changing the rule is easy enough, but once the rule is changed, the hard work of testing and approval has just begun. We need to help our business users through the entire lifecycle of change.
3. The future of rules. The first future is in the past: the marriage of rules and events. This ceremony was performed last year with version 7.5 of Operational Decision Management. The market has yet to fully understand the power of this, but we're hopefully providing the right education. The second future of rules is mobile and, more generally, rich client-side interfaces everywhere. I think we're just entering a "configurable world” where 1990's concepts of mass customization will be applied to everything, everywhere. Customize your shipping options? Customize your notifications? Customize your banking products?
All of these ideas are being worked on, and they all require rules. We're moving from a world where rules engines are primarily bought for managing a few, complicated rule sets to a world where the issue will be managing millions of fairly simple rule sets. We need an engine that scales in that world, a management platform that provides governance and visibility in that world, and beautiful software that enables that world.
Q. Any final comments, predictions, birthday wishes?
Phil: I'd like to thank all the people from ILOG -- those still at IBM and the people who have left over the years, and their families. I know what it takes to have a dream and work to fulfill it. It doesn't happen easily, and it's rarer still to achieve that dream.
Pierre: ILOG has been an incredible experience because we were such a small team fighting against much larger competitors. I’m truly thankful to our customers who took a lot of risks by trying a new technology from a new company. Their success over the years was our success, and was the proof that we were headed in the right direction. They gave us a profound sense of purpose. I think the next 25 years will see the deployment of these concepts across the world and will provide another type of equally exciting adventure, filled with as many surprises as the first 25 years.
Happy Birthday, Business Rules. And many more.