Business Intelligence Bootcamp Part 3

May 17, 2008

We finished up the BI Bootcamp yesterday, and after five days I think I’ve done the Matrix Brainload thing, and absorbed a whole stack of information. This was definitely worth it.

I now understand the basics for constructing OLAP cubes, what kinds of queries I can make to obtain existing statistics, and also to support budgetting and forecasting. I know how to work with dimensions and facts, constructed some MDX queries, created some KPIs, and also covered performance and partitions.

I also learnt how to access the OLAP cubes from SSRS and Excel. With Excel, you link OLAP using Pivot tables, which every good financial controller knows how to do. We also used Microsoft’s ProClarity product to do advanced data analysis. The area that I was most interested in, though, was BI’s predictive ability. We were introduced to the various predictive algorithms, and did some queries that involved probabilities and confidence using the OLAP cube data.

Then we built Sharepoint Dashboards to enable us to represent the information that we built in the labs. This was great, because not only did we represent the data from the cube as figures, but also in graphical form and in a way that enabled us to visually identify the implications of what that data was telling us.

By implications, I mean that you can compare something like Sales Goals to Actuals, you could draw a bar graph in green if the goal was met, or in red if it didn’t. If it didn’t meet the goals, you were able to drill down the hierarchy to identify where something went wrong, or if someone was exceptional, drill down and see if you can identify the secret of their success! Of course, most of the graphing ability has been around for a while, but by putting it in Sharepoint, you are making it potentially accessible to a much wider audience.

We did Performance Point Scorecards, setting a whole lot of options, and were able to analyse the performance using Performance Point Server. We were able to select Analytic templates to do charting, grids, etc. We also built a Performance Point dashboard.

Finally, we did a module on planning, budgetting and forecasting. This was probably the biggest module and included the most significant lab exercise of the entire course! The outcome was that we were able to produce Sales Quota Forecasts.

The course has ended and I’m back in Melbourne. Would I do it again? I think I probably would. Peter Myers is very articulate, and explained key points well. To be able to pick up so much in such a short period of time is just so beneficial. I benefit because I now have a new set of skills for my repertoire, and businesses benefit because, with a small set of skilled people in this area, they are more likely to be able to find someone to deliver their intelligence solution! I highly recommend this course.

Business Intelligence Bootcamp Part 2

May 14, 2008

Well, it’s certainly been an enjoyable ride. I never realised that business intelligence actually covers Sql Server Integration Services (SSIS), Sql Server Reporting Services (SSRS) and Sql Server Analysis Services (SSAS). The BI practitioner needs all three to be effective. I had a lot of previous experience with DTS packages, but it has evolved considerably and is miles ahead of its predecessor. I had also done SSRS but even with significant knowledge, there’s still stuff to learn. And Analysis Services, well, the last time I touched an OLAP cube was around 1999, so it’s a life time in computer terms!

Yesterday we did a whole lot of SSRS to support Business Intelligence. I had done a lot of SSRS before, however I would never make the claim that I know everything, and so I made sure I covered all the steps in the labs, and I was glad to do so, as I did learn different ways of doing things. One of the major things I learnt here is about Report Builder, which is self-service reporting. It’s sort of like a reduced feature report designer that is targetted at Business types that don’t want to go through developers just to get reports done. That’s pretty neat, because what developer really wants to spend their life doing reports! The course also covered integrating the reports with .net applications and how to do it programmatically.

And finally today we got onto UDM (Unified Data Model), OLAP, building cubes, working with dimensions and facts, calculations, KPIs, processing and deployment. I think so far, the most enjoyable thing for me was the SSIS, because I learnt so much, and the topic that required the most concentration is most definitely the construction of the data cubes.

Half way through the course now, and it’s pretty full-on. I have to say if you’ve been around IT for a few years and know your stuff (say, intermediate level and above), and you’re looking for a challenge in an area you don’t know that much about, then BI is definitely a choice worth looking at, and an intensive course will help you hit the ground running as quickly as possible.

Business Intelligence Bootcamp Part 1

May 13, 2008

I decided to do a course. Basically, I decided that rather than fumble through books and notes and help files, I’d actually go and learn a whole stack of new skills and upgrade some old ones. Courses are great, because you can learn an awful lot in a very short period of time, and if you choose the right one, you can learn best practice and learn it from the best.

As my strongest skill is actually in SQL Server, I decided to seek out something that would extend that. I decided that because database, and data concepts are used extensively in BI, it would be the logical choice.

For someone with as many years experience as I have, it is often difficult to find courses of a suitably advanced level, that will also help you hit the ground running in whatever the chosen skill is.

Well, I found the course of choice at Greg Low’s new company SolidQ. The course is Peter Myers BI Bootcamp. See Peter Myers is one of the top BI experts, if not the top BI expert in Australia. He wrote the course!

Anyway, I’ve now completed Day 1, now into Day 2. It’s actually been quite mind broadening. Basically, I had no idea exactly how extensive the need for Sql Server Integration Services (SSIS) is in Business Intelligence. To become an expert in BI, you will need to spend a lot of time taking the data from your data source, massaging it, and placing it in Fact and Dimension tables in your Data Warehouse database. And SSIS is just so rich! There are so many controls in the toolbox that can be added to the design area that it’s simply brilliant.