Netezza’s Enzee Universe conference is now almost over, and I still haven’t figured out what my gig as “conference blogger” entails. More precisely, I’m operating from our unspoken fallback plan, namely “If all else fails, do what you’d do anyway, but do more of it.” For me to live up to that, all Netezza had to do was find interesting things to write about — and as far as I’m concerned, they already did that last Thursday in spades; the five interesting meetings they set up for with users and partners on Tuesday were just gravy.
Another part of the deal was that I’d give a talk this morning at 9:30 am. And when I give talks, I like to put up posts that cover whatever material I haven’t written up before, while also offering the talk’s listeners convenient links to materials I have already covered previously at length.
As I’ve been doing all year, I plan to start the talk with the subject of liberty and privacy. My most recent overview post on privacy and liberty has a bunch of links to what I and other people have said before.
This year’s Enzee Universe keynote speaker was Stephen Baker, author of Numerati. His talk, the only one all week I’ve attended in its entirety (I do intend for mine to be the second one, however ) reminded me that some of my ideas had been inspired by his book, specifically the part about sensors in our elderly relatives’ homes tracking every movement, for legitimate reasons of health care and physical safety.
One thing I’m reminded of when talking with users is that they tend to be a bit focused on their projects or areas, and almost never have the opportunity to consider the full range of possibilities open to them. So I’ve put in two slides to raise consciousness on the point.
- $ per user
- $1000s or maybe $10,000s (perhaps a small team of analysts looking at Big Data)
- $100s or maybe $1000s (perhaps conventional BI)
- On the order of $.10 or $1 (stakeholder-facing analytics)
- Three benefits of better price/performance
- Do the same thing, cheaper
- Do the same thing, better
- Do something different
Then I put together a list of “cool technologies in analytics” people might want to think about, including:
- Solid-state memory (there’s a whole section here about that; see the sidebar)
- Data mart spin-out
- Exploratory BI (e.g., QlikView)
- Advanced analytics (platforms)
- DBMS-centric (more on that coming soon, but meanwhile February’s posts can be a placeholder)
- MapReduce-centric (there’s a whole section here on that too)
- Advanced analytics (UIs and algorithms)
- SQL tasting (I just coined that to talk about the useful idea of getting fast, partial results on long queries)
- Stats/predictive (I’ve really got to build a blog category for that)
And finally, I listed three “aggravating analytic challenges” in areas where I’m disappointed with the progress of and/or prospects for technology, including:
- KPI management
- Text/tabular integration
- Profile of Revealed Preferences aka “Social graph”
Posts on Netezza’s announcements around the time of Enzee Universe
- A long discussion of Netezza’s technology, focusing on the database parts
- A discussion of Netezza’s and IBM’s compression strategies
- Notes on how Netezza balances its silicon and uses its FPGAs
- A quickie on data warehouse loading latency
- How Netezza Migrator works
- Netezza’s strategy for RAM and Flash