Know Your Data Architect

Data Architects are a weird bunch. We love being the man (/woman) behind the curtain that makes the whole enterprise run.  Those front end guys may do all the pretty work, but they will soon lose their golden boy status if they crash our database. Some of us even have a sense of humor, but be careful testing out that theory...

After being in the midst of data architects for many years, there are definitely some stereotypes (myself included!). So here is my poor attempt at understanding architects.

1. High Priest. Tends to be imperious. Feels conflicted about DBAs - loves to spend time with them, but really considers them a lower species. Requires adoration from all other IT staff. Last time they were wrong was in 1965 (when they wore mismatched socks). A traditionalist in data modeling. Anything under a hundred tables is barely worth attention. Probably an Oracle acolyte. Started with SQL Server, but now considers it a hobbyist database. Favorite drink - coffee or dark ale. Favorite food - prefers to eat alone, rarely seen in public consuming food (that tends to enhance the myth). Favorite band - favors obscure alternate bands (especially from eastern bloc countries). 

2. Colonel Clink. Defined the database standards in 1985 for SQL Server. Lives by the same rules in 2015 for all databases. Believes in absolute attention to detail.  No creativity is supported. Abstraction is not allowed. Lookup tables required, Foreign keys required. Requires a mainframe for any database model to achieve any performance. Will only mentor junior architects with the same philosophy. DB2 is the database of choice, but starting to see Oracle as a viable alternative. Can be hostile if you suggest database changes or removal of foreign keys. Favorite food - only eats with their clan. Favorite drink - definitely water. Favorite band - will say Rolling Stones, but really a Journey/Boston/Air Supply fan.

3. Zen Master. Open to all philosophies. Achieves a state of zen with each data model. Feels the data flow through the model. Each table pulsates in color. Managed to design a data warehouse with two tables and one analytic. Everything can be encoded for absolute efficiency. Tends to be hard to communicate with - as they think in six dimensions at any one time. Business users love them, but don't understand why. Other data architects love working with them - they continuously push the envelope, never a boring day. Favorite database - Netezza (of course!). Favorite food - whatever is on hand. Favorite drink - whatever you buy them. Favorite band - one of 1,000 bands. It depends on their mood.  

4. Anxious Arthur. Took a 2 day class in 1972. Living on the edge since then. Trying to stay current, but feels hopelessly inadequate most of the time. Masks this by building highly detailed documents, and taking thousands of micro steps to build your model. An expert in obfuscation. Your model will likely be delayed by a tertiary shift in the tuples that causes a re-examination of the data sources. Worries constantly that the database will fall on its face. Tries to be best friends with DBAs - as they will keep the database from falling on its face and if they are friends will be hesitant to reveal any embarrassing failings in the database. Favorite database - SQL Server (Oracle is too scary, Netezza is too weird). Favorite food - who knows what is in the tupperware containers. Favorite drink - tea (had to give up coffee). Favorite band - too busy to listen to music.

I am sure there are more types. What makes me smile - is that most of us have characteristics from each of these types (and maybe more than one type at some times). It can be both fun and funny to be a database geek (my friend Andre Byfield will vouch for that).