IOD Conference Day 2: Quick comments

This was the first public day of the conference. Just a small gathering of over 7,000 people.....

General Session: Dana Carvey was the master of ceremonies...he was great. It's a good thing, too, because with the 8am start a lot of us were only partly awake. Anything less interesting and we might have nodded off.

Two product announcements caught my attention:

• Data Studio -- to manage the data lifecycle. It's described as the equivalent of Rational, but for data. It's eclipse-based, and we're apparently going to release it as a free download later this month. I'll be at the front of the queue.
• Information Server Fast Track beta -- Doesn't launch completely until today. I'll come back to this in the next posting. It sounds as though it'll speed data etl/mapping tasks. I'm hoping there'll be a data quality aspect to it as well. Perhaps some new/improved QualityStage rule sets?

I attended a banking CDI (customer data integration) presentation by Fifth Third Bank. One of their "lessons learned" was to profile the data before getting very far into the project. I couldn't agree more. There's ample statistics that pre-profiling the data improves project estimates, reduces delays, and saves money.

[soapbox on] I once heard a (rather senior) consultant say it wasn't necessary to profile the data for an SAP implementation, as the application would kick out bad data during the load. [pause for a moment and let that one sink in....] Umm...no! The app will kick out records that don't match its check tables. Unfortunately that's just one small part of the overall data set. Problems with data quality, standardization, and harmonization are a large part of what puts those big projects into delays. [climbing off my soapbox now]

Most of the afternoon was spent in meetings. I did have a good conversation in the Expo with the Princeton Softech guys. DBAs, this stuff looks cool.

Question for the DBA community: Any of you starting to wonder if you need to tool up on the management of unstructured data? As I walked around the expo, I was surprised at the number of vendors and services springing up around unstructured data. I've started to wonder if I need to expand my skillset.

Followup question: Given the known industry challenges managing quality in structured data, what do you forsee enterprises doing to ensure quality and governance in their unstructured content?