I'm about to attend day two of Teradata's annual third-party industry influencers meeting. I'm grateful for the opportunity to listen, to voice my opinions, and, per the agenda distributed by analyst-relations manager extraordinaire Kim Dossey, to "Ask Any Question." I'll actually aim for seven. Here they are...
I'm about to attend day two of Teradata's annual third-party industry influencers meeting. I know of no company that makes a more serious and sincere effort to reach out to independent industry analysts (who follow enterprise data warehouse (EDW) vendors), not only as a market-messaging channel but also to hear our perceptions of their and their rivals' technology, solutions, and positioning. So I'm grateful for the opportunity to listen, to voice my opinions, and, per the agenda distributed by analyst-relations manager extraordinaire Kim Dossey, to "Ask Any Question." An invitation like that one doesn't come along too often from a company of Teradata's stature. Nonetheless, I'll take liberties and assume that "Ask Any Question" means I can attempt to get in seven. Here they are and, noting that I'm probing sensitive and proprietary areas with some of them, I'll plan to post what I learn to the extent I'm able.If you read my stuff regularly, you know that I focus on domains that push up against the boundaries of established BI and other analytical technologies. Teradata is well positioned in some of them including "big data," database parallelization, and embedded, in-database analytics. (Disclosure: Teradata has paid me to write an in-database analytics white paper, which should appear soon, and a couple of customer case studies, and the company paid for my flights and lodging for the influencers meeting.) So by way of "Ask Any Question" questions, let's go for edge areas that I see as important to analytics users and therefore to Teradata, but where Teradata's solutions are perhaps not fully realized.
Complex Event Processing (CEP) is a set of technologies for continuous analysis and transformation of high-velocity, high-volume "data in flight." The most significant use cases are for automated financial-markets trading, yet there are applications that related to enterprise data warehousing -- on-line commerce, communications monitoring, logistics -- are emerging and growing in importance. Even Microsoft has come out with a CEP offering, SQL Server 2008 Release 2's StreamInsight, albeit (it appears) with analytical latency of around 5 seconds rather than the millisecond-level latencies of platforms such as Streambase's; IBM, Oracle, and Sybase have non-BI CEP options whose performance is similar to Streambase's. Yet as of mid-2009, the Teradata platform's minimum latency was over 20 seconds. What are Teradata's thoughts and plans regarding CEP?
It's a truism that approximately 80% of enterprise information is in unstructured sources, primarily textual. Teradata has a long-established partnership with text-analytics vendor Attensity and has recently started working with other text-analytics companies. (Teradata also a long-standing, extensive, and close business and technical alliance with SAS, which has very strong text-analytics capabilities, although Attensity does seem to be Teradata's text-analytics partner of choice.) Yet hosting text processing on the Teradata platform falls far short in business utility compared to integrated analysis of text and data, even more to moving text processing in-database. What is the current state of Teradata text analytics, and what are Teradata's plans to more tightly integrated text into the company's enterprise analytics platform?
2010 is the Year of the Graph, the graph in this case being a non-relational model for network-structured data including social networks and the Semantic Web's aspired-to Web of Data. I know that the Teradata platform has the ability to manage non-relational XML data objects. How extensive are Teradata's XML storage and query capabilities and, more particularly, through XML or through another implementation, to store and support query of graph-structured data?
Three technical questions are enough. Let's move to one half-technical, half-business question, followed by two business questions, and a final dual question:
A number of Teradata's analytical DBMS rivals offer software EDW solutions that run on commodity hardware. Surely over the years Teradata has considered untethering software from hardware. What are your thoughts on this topic, on the feasibility and desirability of a hardware-software unbundling?
What analytical DBMS rivals do you encounter most frequently in competitive sales situations, and how often do you see them?
What factors provided the edge in your competitive wins, and what sales have you lost to rivals and why?
What do you see as Teradata's Top 3 market opportunities -- customer or technology related -- and what do you see as Teradata's Top 3 market threats?
I have no illusion that I'll get responses I can post to some of these questions, especially 5 and 6, or even that I'll get much at all by way of response, but that doesn't mean I can't ask. (Or rather, try to ask: it's not as if I'll be the only one there with questions.) Teradata has few serious rivals in the enterprise data warehousing market, and the invitation to "Ask Any Question" is just too good to pass up. I'll post what I can to Twitter (@SethGrimes) with the hashtag #TD3pi and perhaps in a subsequent blog article.
Addendum, April 28, 4:48 pm Pacific
I got in three of my questions, the three technical questions. Here are Teradata CTO Stephen Brobst's answers, as tweeted by me, numbers matching the question numbers above:
MapReduce is, of course, an approach to parallel processing that is implemented, notably, by Apache Hadoop and also as a parallel database programming framework by Teradata competitor Aster Data. Teradata is pursuing integration of external Hadoop resources into the Teradata DBMS engine.I'm about to attend day two of Teradata's annual third-party industry influencers meeting. I'm grateful for the opportunity to listen, to voice my opinions, and, per the agenda distributed by analyst-relations manager extraordinaire Kim Dossey, to "Ask Any Question." I'll actually aim for seven. Here they are...
The Agile ArchiveWhen it comes to managing data, donít look at backup and archiving systems as burdens and cost centers. A well-designed archive can enhance data protection and restores, ease search and e-discovery efforts, and save money by intelligently moving data from expensive primary storage systems.
2014 Analytics, BI, and Information Management SurveyITís tried for years to simplify data analytics and business intelligence efforts. Have visual analysis tools and Hadoop and NoSQL databases helped? Respondents to our 2014 InformationWeek Analytics, Business Intelligence, and Information Management Survey have a mixed outlook.