Last year I followed up the TDWI Conference in Orlando with a kayak trip to a few of the swamps and rivers of central Florida. That trip proved to be rather glorious, and so my friends and I
decided to tie the two events together again this year.
The kayaking was better than ever, but more on that later.
Sunday
I rolled into Orlando Sunday afternoon and headed for Bindu Shah’s night school session, “Data Warehouse Appliances – Too Good to be True?”. Shah reviewed two vendors, Netezza and DATAllegro.
Both vendors combine server hardware, operating system, database software, and storage into single systems that are tuned to support high performance query and load operations. Because all of these
elements are preconfigured, data warehousing appliances offer rapid deployment. However, Shah cautioned that these systems are not for mixed use, i.e. BI analysis and transaction processing
support. She also suggested that buyers include concurrent user load testing in their proof-of-concept evaluations.
After night school, I stopped by TDWI’s Reception to network and cruised by the sponsors for the evening — Hyperion, Identity Systems, Informatica, Initiate Systems, Relational Solutions,
Teradata and SGI. I had worked with SGI in the past, and although they were known primarily for their high-end work stations, I had seen their servers used for serious data analysis. It was good to
see SGI at a TDWI conference.
That evening I had dinner with friends at a superb restaurant, which was in the conference hotel – the Royal Pacific Resort at Universal. Emeril’s Tchoup Chop featured Asian and Polynesian cuisine
put together under the supervision of Chef Emeril Lagasse. The food was great, the décor was inviting, and it was a great place to reconnect with old friends (and young ones as well.)
Monday
Jill Dyche’ got everybody’s attention with her keynote, “Your Data Warehouse is Dead! The Rise of CDI,” which was a pretty gutsy title for a data warehousing conference. Actually, Dyche’
wasn’t suggesting that data warehousing was no longer relevant. To the contrary, she pointed out that the data integration skills learned by data warehousing professionals can be leveraged in CDI.
CDI stands for customer data integration and is a part of a master data management environment. Dyche’ explained that unlike the data warehouse, CDI “supports up-to-the-minute provisioning of
customer details to other application systems.” She was also quick to point out that the CDI and the data warehouse serve two distinct functions. The CDI integrates data at the operational level,
and the data warehouse integrates data for historical analysis.
Having a keynote about data integration issues provided a nice mental warm-up for Evan Levy’s session, “Beyond the DW: Architectural Options for Data Integration.” Levy started off with
connecting the dots for us as to why data integration is more important now than it ever has been. He cited several cases of companies incurring fines for violating do-not-call statutes or
soliciting customers from areas where they did not have rights to do so. He also listed several enterprise data integration myths, i.e. …integration means purchasing an ERP or integration means
implementing a data warehouse. While these activities require data integration, they don’t address the complete integration needs of the typical organization. Levy then went on to identify various
data integration drivers, such as operational customer relations and data analysis – and the strategies that companies can use to support them.
Levy spent the rest of the session covering the differences between ETL (extract, transform, and load), EAI (enterprise application integration), EII (enterprise information integration), and CDI
(customer data integration). Most data warehousing teams rely on ETL strategies for their data integration needs. However, some teams that are involved with real-time data warehousing are using EAI
technologies for their data feeds. EAI usually requires program modification of source system programs or installation of listener programs on those systems. However, the result is that the data
warehouse can potentially reflect all changes that occur in those operational systems. EII usually involves a central server that provides SQL based query access to all sorts of operational data in
various formats via data adapters. Levy described CDI as “…the collection of processes, controls, automation, and skills necessary to standardize and integrate customer data originating from
different sources.” As such CDI might take advantage of EAI technologies. By the end of the day, my classmates and I had a pretty good high-level grasp on these technologies and strategies.
That evening, I was ready to kick back and relax. Fortunately Oracle and HP hosted a mellow hospitality suite with good food and Jeff Scott who played exquisite classical guitar with a Spanish
touch. It was perfect place to relax and network after a long day.
Tuesday
Given that the value of a data warehouse is only as good as the usefulness of the data it contains, TDWI has been advocating that professionals in the field become familiar with performance
management and its supporting applications – performance dashboards. Wayne Eckerson, Director of Research and Services at TDWI and author of a book plus a TDWI study on the subject, was presenting
a session titled “Performance Dashboards: Measuring, Monitoring, and Managing Your Business,” so I took the opportunity to sit in on his session. Eckerson covered the history and reasons that
organizations have adopted dashboards. He included a number of short case studies, which always makes a subject come alive. He also pointed out that the data warehouse provides the natural supply
data for a strong dashboard program that includes heavy-duty analytics. In other words, if a key indicator is out of whack, can I quickly ascertain the factors that contributed to the situation?
Eckerson went on to cover several different architectures, 15 required features, and deployment strategies. In the end, I felt that the day was well spent.
The vendor show was in full swing during the extended lunch break and I took the opportunity to visit three of the major database vendors in the data warehousing space, Teradata, IBM, and Oracle.
Our industry has asked these vendors to support larger and larger data volumes while providing faster and faster access to whatever data we might need. Teradata has had a long association with data
warehousing, and while the Oracle and IBM general marketing programs might not stress such support, each one of them continues to make significant strides in the data warehousing space. I was
impressed by the depth of knowledge each one of these vendors had in their respective booths. Anybody having questions regarding data warehousing applications would have had little trouble in
getting their questions answered.
That evening as I walked down the hall, I met two show girls who introduced me to Elvis and invited me in to the Sun Microsystems/ Informatica/ GreenPlum/ AMD hospitality suite. Inside Viva Las
Vegas was in full swing with gaming tables, a free stack of chips, munchies, and a well stocked bar. What a writer has to go through to cover TDWI!
Wednesday
Several years ago, I had the pleasure of spending some time with Arkady Maydanchik, who was one of the more insightful folks I had ever met. I remember a long after-dinner conversation between Dave
Wells and Maydanchik about the issues surrounding data quality. When Wells, who puts together TDWI programs, scheduled Maydanchik for a session, I jumped at the chance to attend his “Practical
Skills for Data Quality Assessment.” Maydanchik didn’t disappoint. He delivered one of the most thorough overviews on data quality that I have ever heard. Towards the beginning he encouraged
organizations to create a data quality/data stewardship group that spanned IT and business. Maydanchik’s session provided a sound level-set of knowledge about data quality assessment issues for
everybody in the group, but it also went on to challenge most of us who thought we had a fairly good handle on the subject. Those without a background in data might have found it a bit heavy-going,
but from my vantage point, the challenge was worth it. My guess is that some of my classmates from the business side of the aisle, headed back to their organizations with a few insights that might
challenge their IT brothers and sisters.
That afternoon, I sat in on Cindy Howson’s “Evaluating BI Toolsets.” Howson offers this course quite often. However, she works on keeping it up to date, so it is worth taking every-so-often just
to keep up on BI technology. She also schedules different vendors to go head-to-head on a tough set of BI challenges. This time, Microsoft, Oracle, and Information Builders took the hot-seats.
Microsoft’s foray into the BI toolset space has had pure-play BI vendors looking over their shoulders, and other database vendors re-examining their BI strategies. However, during this shoot-out,
I thought that Oracle and Information Builders came off as having much more mature solutions. I was particularly impressed on the depth of functionality that Information Builders demonstrated, i.e.
their data analysis/visualization option, Visual Discovery, and their portable analytics, Active Reports. I came away thinking that Microsoft still has a ways to go and that any organization
looking for a BI toolset should still consider the pure-play BI vendors, including Information Builders.
Thursday
Claudia Imhoff has written or coauthored a number of books on data warehousing and related topics. She continues to contribute to the field through online publications and her blog. She speaks
frequently at TDWI and other conferences. When Dave Wells introduced her, he said that he had googled her and came up with 50,000 + hits. Although Imhoff is definitely one of the thought leaders in
BI and data warehousing, I wasn’t particularly looking forward to her keynote, sternly titled, “The First Word in BI is Business.” I had seen a lot of lectures about dysfunctional business/IT
relations, and wasn’t looking forward to yet another. However, Imhoff surprised me. In forty minutes or so, she had called out most of our more paternalistic excuses for not cooperating with the
other side, told us how to spot attitude problems on either side, suggested that we walk from projects that don’t have active business participation, and given us common ground on which to build
better relationships and more effective BI systems. I was impressed.
When Imhoff had finished, Dave Wells honored her contributions to the discipline by making her a TDWI Fellow. Fortunately, Wells did wait until she had delivered her keynote, because she was a bit
speechless afterward. Nobody has been more deserving.
After the keynote, I headed off to Michael Gonzales’s “HandsOn-Risk Mitigation for Business Intelligence.” Gonzales is a great presenter and his “HandOn” series of classes is usually hard to
get into. Fortunately, I had registered early.
Gonzales went over several techniques for reducing risk for BI projects, including business rule audits, proof-of-concepts, and data profiling. He also talked about Larry Boehm’s Spiral Approach
to project management which consists of four quadrants: determine objectives and constraints, risk analysis and alternatives, development, and planning for the next phase. As promised, a good piece
of the session was devoted to hands-on work where we worked through problems that included how to do business rule audits with Syncsort’s DMExpress.
Friday
I caught an early flight out in order to make my next gig, but once I settled into my seat, I started thinking about the next TDWI Conference in Las Vegas in February. It should be a good one. See
you there.
Post script
TDWI weeks are jammed packed with content. It’s always wise to schedule some down-time before or after the conference. Due our schedules, my friends and I decided to go kayaking prior to the
conference. We established a base at the Crystal River Resort in Crystal River, Florida. From there we explored the Chassahowitzka, Santa Fe, Homosassa, and Crystal Rivers over the next three days.
The slow-moving creeks leading into the Chassahowitzka provided the classic Florida swamp experience. We paddled into back water springs and enjoyed some quiet time away from the press of
technology. The next day we headed down the Santa Fe for a different sort of experience.
The Santa Fe ran a bit faster than the Chassahowitzka. We paddled through a few class I rapids and took a few short breaks at some picturesque springs. We often found ourselves in oak forests. Box
turtles and birds abounded.
The next morning we took an air boat trip up the Homosassa and encountered a few manatees. However, by that afternoon, the water temperature had dropped in the ocean and the manatees migrated en
masse up rivers to the springs that offered warmer water. By then, we were out on the Crystal and back in kayaks. We ended up at a spring with over 20 manatees hanging out. They are really mellow
animals and are very friendly. It was a perfect opportunity to chill out and get ready for the conference.
