Via SoCalDevGirl, I’ve just discovered another interesting piece in the somewhat fragmented Microsoft Business Intelligence story: Microsoft Solver Foundation. Here’s the official website:
http://www.solverfoundation.com/Default.aspx
What is it then? Ahem, well, if you can’t make much sense of the blurb on the website (like me) it’s probably not aimed at you. Here’s a sample quote:
Solver Foundation is a Microsoft framework designed to deliver critical business insight tools to CxOs, quantitative analysts and developers of mission-critical systems. Traditionally referred to as mathematical programming, these tools provide business intelligence and planning support to organizations seeking maximal competitive advantage.
I suggest you read the full overview to get a better idea of what it does. What I do understand, though, is that anyone who uses this is going to be interested in using the rest of the Microsoft BI stack; I sincerely hope that the Solver Foundation team is talking to the other BI teams and that some kind of coherent BI strategy will emerge. If one does it’s clearly going to be Excel-centric (which makes a lot of sense): Solver Foundation has an Excel addin; there’s also the data mining addin; there’s SSAS’s own integration with Excel; and of course Gemini will be surfaced through Excel, tying up SSAS, some kind of data cleansing functionality, and possibly some data mining functionality too into one compelling package.
Very interesting! Microsoft is shifting more and more from just a technology provider into business area. When I say technology provider, I mean OS (Windows), developers infrastructure (Visual Studio, etc) and ISW tools (Office suite), things that Microsoft is famous for.People that excel in using those systems (configuring them, etc) are mostly IT people, in other words, people that finished tech studies. Business area is relatively new on Microsoft\’s horizon: the Dynamics suite as Microsoft\’s ERP system, KPIs in SSAS and Performance Point as Balance Scorecard system and now this. In order to be good in them, one should have a solid background in business and economy studies (among the rest!). The shift (or shall we call it attack) is slow and appears from different sides. What\’s common to these new areas is that although they are aimed at highly specialized professionals/consultants, they require a solid technical basis, which makes them a perfect bridge to this new business world. Together with the new SaaS model to deliver them.How I see it is this: any kind of specialized knowledge will be your ticket to success in future because you\’ll be able to deliver it as a service to a wide variety of users needing it; on the other hand, Microsoft will establish himself as a central figure by providing this framework. Just like with OS in old days. Only now we will have services. And this new framework provides endless possibilities. The phrase "knowledge is power" is about to come true in a very concrete manifestation in the following years. While it was always so, with SaaS it will have the long-tail effect boosting it.Enough. If you want to find out more about the "Solver" Lynn mentioned in her blog, this topic ("mathematical programming", as they put it) is better known as "Operational Research" (from what I read about it and concluded from my OR studies). I checked and now the study changed the name to Quantitive Analysis). So I think we should search under both terms. You know how it goes, names get fancy over time.Users of this framework will be solving problems like these: transport problems (finding the optimal route to deliver the goods among N places), blanket problem (how to cut the material in order to have minimal losses), optimal schedule, allocating work on several machines and various other industrial issues that in their background use complex mathematical methods: linear programming, non-linear programming, discrete programming, multi-criteria decisions, regressions, etc. That\’s why the name Solver is adequate, right?By its complexity, these systems are equal heavy to grasp (in their core, not usage) as DataMining is. For geeks 🙂 I might have a peek into this on MSDN these days, it looks like a promising field and moreover, it features problem-solution approach, so tempting to me. I heard stories like people getting nice percentage of savings from implementing some optimization technique to a business (namely banks). Only this time it is very, very measureable, as opposed to some other areas of consultancy. Why? Companies spend money on a daily basis. If you can save them part of it by explicitely pointing a way to do it, next month they\’ll spend less. The difference is known exactly – less gas on trucks, less waste in production, better break-even points, etc.Did you know ABC analysis is one of the things solved in OR? And we had that in our OLAP client way back in year 2000, among other 4 analysis. Too early, don\’t you think :-)?Vendors (I guess ones being attacked by this move are): Aris, Ventana Research, … Or not?A good place to start finding more about this topic is this: http://en.wikipedia.org/wiki/Operational_researchA book used in seminar I had (its current version, of course):http://www.amazon.com/Operations-Management-Strategy-Analysis-6th/dp/0201615452