Enterprise Decision Management - a Weblog

-- Posted by Carole-Ann

Thank you for attending the Evolution of BRMS session at Business Rules Forum.  It was great meeting some of you there in person.

For those of you that could not make it, I wanted to give you the gist of what I presented.  A previous presentation covered the evolution of the business rules technology focusing first on the drivers that forced the market to shift its focus from Business Rules Engines (BRE) to Business Rules Management Systems (BRMS).  In this second presentation, I explore the evolution that is taking place as we speak, going from BRMS to Decision Management (DM). 

In a nutshell, the main ideas are summarized below. 

Increase Confidence in Strategy Performance

Address more Sophisticated Decisions

Connect Decisions

The slide deck is available from the Business Rules Forum website if you are an attendee.  Our marketing guys will also post it on our community site for your convenience.

-- Posted by Carole-Ann

After Dallas, travel brought me to Florida where I attended Business Rules Forum.

On Tuesday, John Rymer invited me to sit in a vendor panel (the only woman).  We discussed topics such as market consolidation and evolution towards Decision Management.  As you would expect all the vendors were unanimous that they would still be in business despite the Platform play of the IBM, Oracle etc. ;-)

I was not surprised to see that BRMS vendors are trying to confuse the audience regarding Decision Management.  As an early pioneer, I am of course a great proponent of Decision Management.  I contributed to shape its vision.  Therefore it is sometimes frustrating to hear people say it is the same as BRMS.  I love BRMS too but I see a clear distinction between the two: one is a superset of the other.  BRMS focuses on business rules, in other words: "how do we empower business users?".  Decision Management looks at the bigger picture in terms of including a larger set of technologies to solve more than business rules problems: "how do we predict the ability to repay?  how do we optimize resource allocation?"  Decision Management takes a serious look at how we improve decision too: it's about providing tools and services for business users to assess how profitably new strategies will perform in various economic conditions.  This goes well beyond business rules validation as provided by the BRMS today.

Frustration can turn in great satisfaction.  Although vendors faked they did not get it (or maybe they really did not), some bloggers did not get it either, I was extremely pleased to see that the audience was getting it.  I heard several people talk about "automating and improving decisions".  They are (getting) ready for this new discipline.  I definitely perceived from the side discussions I had after the panel and my talk that enterprises are now ready to investigate or embrace Decision Management.  I did not get that feeling in the previous years so I am sensing some market maturity there.

The panel touched also on Standards.  I was probably not popular stating that we are not ready for standards.  Not that they are unimportant of course, they can definitely change the adoption of technology.  But I still feel like we are at the stage where vendors want to define standards for the sake of having standards.  I have not seen any real end-user contribute to OMG or W3C on that subject.  I do not expect them to determine what the standard should be (duh!) but I expect them to dictate what is important to focus on.  Is it all about being able to swap one engine versus another, like a rule language extension of JSR94?  Or is it about swapping BRMS and therefore focusing on HOW the rules are represented (templates, metaphors, etc.) and lifecycle management processes?  Or was JSR94 the only thing end customers really cared about?  I am pretty sure JSR94 is not the solution of course.  But I think we will run into a wall if we try to define a standard for the wrong thing.  Not knowing the objective I could define more than one perfectly engineered wrong design.  After all, this is the core reason we have product managers in the software industry: to link the solution to the root business problem.  If we miss the business problem, the solution will likely miss the point too.  Vendors tend to disagree.  Fine.  The illusion that BRMS can be swapped with no effort is pure marketing today.  What is being discussed in terms of Standard will not allow to make it real though.  No offense to the vendors involved, it is all about business.  With pragmatic people like Changhai being involved in OMG PRR effort, I feel that we have a chance to get real results *if* we focus on what the end-users really need.  This is an open call to real end-users with real business problems to speak up.

On Wednesday morning, I presented my perspective on the market evolution to Decision Management.  This was the part 2 of the pre-show webinar I blogged there.

The talk was much better attended than most of the sessions I went to.  If you were in the audience, thank you for coming.  My objective was to educate the crowd on Decision Automation technologies such as Predictive Analytics and Optimization as well as Decision Improvement concepts.  It feels like it was the right approach as the BRF attendees are not yet fluent with those concepts.  I'll provide more details on the talk in another blog entry.

One of the highlights of the show was our Customer Dinner.  We had a private reception at the Fulton's crab house.  Great food.  I also got the opportunity to spend one on one time with a few customers that I had not met before.

Overall, the show was a good Business Rules Practitioner Reunion as always but it was a disappointment to see it had not grown much.  It is understandable that the attendance is down with the crisis but it is unfortunate that quite a few sessions had to be canceled.

-- Posted by Carole-Ann

The very first occurrence of October Rules Fest happened in Dallas a couple of weeks ago. The objective was to get all kinds of business rules techies in one room and cover all kinds of topics from rule algorithm details and innovation to enterprise deployment considerations to decision management.  James Owen and team assembled a fantastic panel of presenters if I may say so myself.

Despite the slowdown in the economy the attendance was pretty good -- around a 100 or so senior architects and CIOs from large enterprises from all around the world.  Given that they did not do any marketing, this illustrates the power of blogs!

Everyone was very engaged in the sessions.  Lots of great conversations during and after the sessions.  It was friendly, competition was respectful and most of us stayed at the technical level (no sales pitch).

Being one of the early Decision Management pioneers, I decided to provide education on Predictive Analytics.  I changed the title to "Demystifying Predictive Analytics" by the way.  From the feedback I received it seems that it stroke a cord with the audience.  Too often, modelers and IT do not talk (that is if they know each other at all!) and do not value each other's contribution.  My talk intended to provide enough background on what models are, how they are being constructed and what they can be used for.  It took us years at Fair Isaac to get mutual respects.  Sharing our scars will help the industry at large "grow up" faster.

Carlos Serrano-Morales and myself had the opportunity to present again on Thursday afternoon.  We focused this time on Enterprise Deployment.  We admit this is boring stuff to some extent but this is also what business rules evangelists need to understand to really exploit the whole value of business rules, and avoid architecture pitfalls.

Charles Forgy presented also the latest and greatest innovation on parallel Rete.  I can tell that all the vendors were taking notes!  Aren't we lucky to have him working for us?  Priceless.

Again, based on James Owen blog, it seems those three sessions were quite well appreciated by the audience.  I encourage you to read his blog report on the sessions.  The slide decks will be available soon I believe.  They also taped the talks.  I am waiting for the videos any day now.

I loved the show.  I hope that James and Rolando will do it again next year.  It was a great place to discuss the "technical" stuff without the marketing.  It felt like a high school reunion, just missing a couple of the "usual suspects" (Paul Haley and Ernest Friedman-Hill).

Feedback from our customers was also very positive.  So this is not just my opinion...

I also want to thank Andreas for the picture of me presenting.  If you are interested in what I am showing on the screen, it is a pretty cool data discovery report from Model Builder.  I am explaining the weight of evidence sparklines.

-- Posted by Carole-Ann

Being part of Fair Isaac, I have never really doubted the critical role predictive analytics could play in business.  I see real-life examples day in and day out but when I read the New-York Times article http://www.nytimes.com/2008/10/05/business/05fannie.html?partner=permalink&exprod=permalink, I found one that was painfully telling…

Let me point your attention to a few points that CHARLES DUHIGG makes in this investigation.  I will not comment though on the politics involved.

Fannie Mae has been playing a crucial role for the lending industry as we all know.  Their early success relied on their ability to predict which borrowers would be able to repay, assessing the premium required to compensate for the risk they took.  With an effective model, you can pinpoint good risk versus bad risk and therefore make safe decisions that ensure the business will prosper.  This is pretty much the essence of predictive analytics: once you know the probability to repay for people with a given set of characteristics, you can extrapolate and estimate how much reserves you need to build in order to beat the odds that you will not get you money back for each population segment.  This is how premium are calculated.

Unfortunately with new types of mortgage product and the lack of associated data, Fannie has not been able to produce a robust predictive model.  Not being able to tell how borrowers would behave in the long run, there was no way to estimate how much risk they were exposing themselves to.  When you combine that with the absence of a CRO (Chief Risk Officer), which is like driving with your eyes closed, no wonder we got where we are.

This is fairly atypical black and white example but it drives the point: with a robust model, you thrive / without a model, you’ll hit a wall… eventually.  I am not saying that predictive models are the only way to make good decisions but they definitely help manage the uncertainty.

One might wonder how they could have built a robust model in absence of data.  This is a fair question.  Well, how about better integrating a feedback loop into the system to detect the early signs of the meltdown?  This is what we call Decision Improvement.  Another potentially complementary approach might have been to take a less aggressive stand: knowing that we don’t know for sure, premiums could have been calculated with more room for caution.  This is where business rules and predictive models work hand in hand.

Again, this is not taking into consideration other facts that influenced the situation such as the mandates from Capitol Hill or the laxity of credit agents.

-- Posted by Carole-Ann

We were all waiting for Oracle to respond to the IBM acquisition of Ilog.  Here it is.

http://biz.yahoo.com/prnews/081029/aqw128.html

With the Haley acquisition, Oracle joins the BRMS party.  Oddly enough, they went for one of the smaller players.  As you may or may not know, Haley (recently acquired by RuleBurst, previously known as Softlaw) has been focusing on natural language.

They have been mostly successful in Government as far as I know, but we never really competed with them head to head.  This makes me wonder what to expect in terms of positioning...  Are they going to keep focusing on their niche market?  Are they going to invest in the product to make it more "like the others"?  Were is it going to fit?

Now we know why Haley did not show up at Business Rules Forum this year!

-- Posted by Carole-Ann

Thank you for joining the Evolution of BRMS webinar yesterday.  It looks like many of you have been able to make it.  Quite a turn out!

For those of you that could not make it, I wanted to give you the gist of what I presented.  This presentation covers the evolution of the business rules technology focusing first on the drivers that forced the market to shift its focus from Business Rules Engines (BRE) to Business Rules Management Systems (BRMS).  In a nutshell, the main ideas are summarized below.  In a few days, the recording will also be posted on our community site for your convenience.

Empower Business Users

Manage End-to-End Business Rules Lifecycle

Deploy Seamlessly and Consistently throughout the Enterprise

In the second part of this series, I will explore the evolution that is taking place as we speak, going from BRMS to Decision Management (DM).  If you attend the Business Rules Forum in Orlando, I invite you to attend this presentation live.  It will be packed with details on the decisioning ecosystem, including predictive analytics, optimization and key decision improvement technologies and techniques.

-- Posted by Carole-Ann

Interested in a live performance?  In preparation for Business Rules Forum, it will be my pleasure to offer a free webinar on the evolution of the business rules management systems on October 8th.

Details on the Business Rules Forum website, under the "free webinar" category.

-- Posted by Carole-Ann

I remember the days when BRE (business rules engine) was a hot topic.  The analysts tracked the BRE market, customers looked for engine features, vendors were competing against spaghetti code.  Back then we were a handful of passionate individuals and we believed there was more to it than just execution.  We made BRMS happen.

BRMS made it because it addresses real problems: the management of those business rules.  Having a BRE is better than code for sure but without management you only get another technique of writing code.  Management allows you to make your policies visible, auditable, trackable.  With key features such as the Rules Maintenance Application (RMA) and a truckload of repository services such as versioning, verification, validation, and finally lifecycle management, end users are empowered.  They are part of the picture and they are accountable for the decisions companies make.

BRMS is at its top, with now the attention of platform vendors, and again a handful of passionate individuals are putting into words the real problems that still need to be addressed.

A couple of years ago, we saw interest in that vision.  Early adopters jumped into the Decision Management bandwagon.  Now this is getting more and more of a reality, Decision Management is officially there to stay (until we reach the next frontier).  BRMS is not dead, but it is becoming a feature rather than the whole thing.

-- Posted by Carole-Ann

On September 30th, we will launch a Decision Management Tools community.  The objective is to allow developers to interact and learn about the DM Tools products.  Business users may find some useful discussion threads too but keep in mind that the focus is primarily technical.

You will find interesting details on the product (new cool features, best practices...), on the market (customer case studies, events...), and your own discussion topics.  We want to open the dialogue we have one-on-one to a global web platform.

We look forward to this new way of interacting and learning more about each other!

See you there in 5 days...

-- Posted by Carole-Ann

With the new school year starting (and my return from extended vacation), I thought it would be appropriate to review a few highlights of the business rules activity...

In January, Blaze Advisor received the 2008 Best BRMS of the year award from Infoworld -