≡ Menu

Teradata Unifed Data Architecture #TD3PI


Next up is Randy Lea with a discussion of Teradata’s UDA – Unified Data Architecture. The challenges of managing and analyzing data have not really changed in years – people still worry about complexity, volume and ease of analysis just like they always have. Competitive differentiation continues to require management  of data and delivery of effective analytics. What’s changed is the ecosystem  supporting this – from a single big box to a more varied, interconnected ecosystem of pieces. Teradata he says remains focused on delivering a lower total cost of data management and a lower cost of analytic.

While the emergence of new data sources like the internet of things or social data is changing the problem domain rapidly it remains essential that the business can understand sales, marketing, operations, customer experience etc. Being able to bring this data together, model and integrate it, and deliver value from analytics built on this data is still essential. And this tightly coupled, modeled data can be extended and enriched with new data sources, big data etc. Teradata’s UDA is designed to support this range of data types – from tightly coupled and modeled data to loosely coupled or even uncoupled data – so that companies can bring new data sources together for analytics.

Once these different kinds of data are being managed effectively there is still a need to manage analytics effectively. Analytics is also changing with time series,  text and rich media analytics for instance all being added to more traditional analytics. Yet the critical issue for most companies when it comes to analytics is that they don’t have the people they need. So leveraging established skills, like SQL or general business analysis skills, becomes critical for organizations trying to broadly adopt analytics.

Randy identified one of my favorite issues – that companies take data, add analytics and then try to see if those analytics can change decisions and hence business processes. In fact you need to start with your business, identify the decisions that will make a difference and then drive back to analytics and data that will make a difference – begin as I like to say with the decision in mind. Randy feels that too many companies start with the technology and data when they need to start with the business problem to make sure that the #analytics they develop will make a difference.

Teradata’s UDA is designed to address these issues – bringing together the data platform, integrated data warehouse and data discovery platform, supporting a wide variety of data types and sources, delivering a set of analytics, and doing so increasingly in real-time. Teradata believes that only a unified architecture that lets them pull together the various  technologies available will allow them to drive the lowest cost of ownership for data management and analytics.


Comments on this entry are closed.