There’s an interesting blog post by James Tayler over at eBizQ on Business Rules, Decisions and Events that talks about the role of CEP in decision management. I share James’ view that trying to compare and differentiate BPM from CEP is tricky and not worth the effort. So much of the problems and solutions end up being along a continuum with overlapping domains, that trying to draw black and white distinctions ends up being arbitrary. Can CEP technology be used in BPM applications? It often can. Do all BPM applications need CEP? Certainly not.

I think James gets to the essence of how we (and others) are applying CEP, which is to either enable or make better decisions. I like the smart order routing example for a couple of reasons - for one because it’s an example of how CEP is being used for things beyond situation detection - the type of use case that gets the most attention when it comes to CEP. Also because it draws from a lot of the different capabilities of CEP: analyzing events in the context of other events; watching for patterns of events (eg patterns that suggest hidden liquidity); correlation of events (comparing trade activity to order activity in the market); and doing all of this in real-time.

Traditional order routing in the securities market followed relatively simple rules. In fact it could be as simple as “if the order is an equity order for less than n shares, send it to x”. In today’s markets, orders are routed more intelligently, and there is a lot of interest in “smart order routing” which takes into account the state of the markets to make a decision on where to send the order. Now even here, there is a range of complexity that can be factored in. At the simplest level, a “smart” order routing application might simply look at the best bid and offer from multiple market venues, and pick the one with the best price. While some call this smart order routing, it really doesn’t take much in the way of smarts. Where it starts to get more interesting is when you factor in market depth - not just who has the best price, but where is the liquidity and where can you get the best overall price for the quantity you need to move? In fact Aleri’s Market Liquidity Analysis framework, which in implemented on the Aleri CEP platform, uses CEP to consolidate and analyze market liquidity in a fragmented market - where a common use case for the analysis is to intellegently route orders.  Then you can further ratchet up the complexity of the routing decision when you factor in dark liquidity, and use algorithms that look at things like refresh rates to estimate where there may be hidden liquidity and factor this into the routing decision. CEP is well suited to this since it’s all about analyzing events (in this case the order) in the context of other events (activity in the market) to make a decision (where  - and sometimes when - to send the order). It’s this type of real-time decision making where CEP can help, and thus smart order routing is just one of the applications in the capital markets where firms are turning to CEP. Is this a BPM application? I guess it is - though probably not in the conventional sense. But it’s certainly a case where CEP can deliver value in automating - and more importantly optimizing - a business process, where there is a level of complexity in the decision making that is required - in this case the need to take into account numerous external factors that are independent of the event requiring a response. Here CEP is deciding the appropriate response to a single incoming event (rather than a pattern of events) but the nature of the response will vary depending on the external conditions as discerned through previous events.

Tags: BPM, CEP, Complex Event Processing, MLA, smart order routing

This entry was posted on Thursday, August 28th, 2008 at 1:01 pm and is filed under Aleri CEP Blog . You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.