Connections
Services
Release the Flow
The Teradata Aster Analytic Pipeline Discovery sets the stage for uncovering new information.
by Paul Barsch and George Kong
Most data management professionals know that multi-structured data such as Web server logs, social media and sensor data abound in their enterprise. However, they may lack a clear view on how to derive value from it. The volumes of information generated from these sources are often referred to as “big data,” which speaks to its complexity, variety and velocity.
Teradata Aster, with its patented SQL-MapReduce capability, provides an ideal solution for running complex queries on multi-structured data to parse through raw clickstream logs, market basket data, transactional data and more. Analysis of these information types helps enterprises discover hidden patterns, understand customer behaviors, and illuminate customer, supplier and partner networks. The passageway to these revelations starts with the Teradata Aster Analytic Pipeline Discovery service.
The Four-Step Roadmap
The Analytic Pipeline Discovery service helps organizations understand their current capabilities in analyzing multi-structured data and provides a roadmap for investigative analytics, digital marketing optimization, fraud detection and more.
Comprising facilitated workshops and interviews by Teradata Aster analytic scientists and technical consulting experts, the service examines the “pipeline of data” linking source systems, data integration processes, business intelligence (BI) applications and business outcomes. The Analytic Pipeline Discovery service consists of four phases: introduce, assess, design and solution.
1. INTRODUCE
Teradata Aster Center of Innovation consultants collaborate with the client’s IT and business teams in workshops to assess preliminary data points; source systems; extract, transform and load (ETL) servers; data warehouse assets; and BI tools.
This phase is a deep dive into the client’s current state, focusing on business drivers, analytics processing requirements and associated IT challenges. The workshops leverage a detailed questionnaire that is sent to a sponsor within the organization for distribution, as applicable, and is completed and returned to Teradata Aster consultants before the first meeting. This “response package” provides consultants with an introduction to the organization’s existing environment.
2. ASSESS
Data discovery examines and documents information sources and existing databases that may serve as source systems to the Teradata Aster platform. The IT asset review displays various data marts, business process management software and ETL servers. Processes for data source extraction, data cleansing, data collection and warehousing, reporting and analytics, and current business outcomes are shown on “as is” topology maps.
Once the data pipeline is documented and data flows understood, Teradata Aster consultants use a field-developed tool to assess analytical queries across dimensions of business value, query complexity and data set size. The most powerful impact of this phase is the determination of “high value” queries that map to business strategies such as acquiring new customers, improving customer retention, better product and service pricing strategies, and improvements in operational efficiency.
For example, an online retailer wanted to improve market share through new customer acquisition. Applying analytics to this challenge included the development of queries on personalizing recommendations, improving product placement on Web pages, and understanding implications of changes to individual Web pages. These queries were scored against dimensions of business value, complexity and size. The output showed the retailer which query would have the largest immediate impact.
3. DESIGN
A key output from this phase is a conceptual architecture that depicts the “to be” topology map of a data pipeline based on the client’s prioritized list of challenges.
An Analytic Pipeline Discovery engagement for an online media company prioritized “speed and agility” and “scalability” of analytics. Recommendations were offered in advanced data collection and loading techniques to process data quickly for analysis. A scalability review also examined the right design to meet today’s analytic needs and to prepare for tomorrow’s data deluge.
Also produced in this phase are examples that demonstrate how the proposed solution will meet stated business goals. Financial measures such as total cost of ownership (TCO), return on investment (ROI) or internal rate of return (IRR) calculations may be included to quantify costs and benefits. A use case could show how the proposed conceptual architecture would meet a client’s goal of providing a rich self-service analysis environment that delivers data and analysis faster. Those business goals would be associated with quantifiable measures such as cost savings, productivity improvements, higher customer satisfaction scores and revenue enhancements.
4. SOLUTION
The assessment finishes with a stakeholder presentation on Teradata Aster’s understanding of client business requirements and architecture. Teradata Aster consulting teams offer senior executives a diagnosis of current analytic capabilities and an action plan to meet future needs. A bill of materials to deploy the solution is discussed alongside business case results. Also included are recommendations for either moving toward a proof-of-concept stage or full deployment of the proposed solution.
Plan for Success
The Teradata Aster Analytic Pipeline Discovery sets the stage for success by capturing and validating requirements, designing a powerful solution that meets business objectives, and developing a deployment plan to gain big-data insights. After the phases are complete, a series of deployment services are identified. These may be focused on building the best analytic platform to meet customer needs, data modeling and loading, building BI reports, and mentoring for tuning practices to keep the system optimized.
Teradata Aster’s service methodology enables organizations to quickly analyze multi-structured data to uncover patterns, behaviors and networks. It also allows companies to obtain value and insight from previously untapped sources of big data to benefit their business.
Paul Barsch is a marketing director in Teradata Professional Services and has more than 12 years of IT experience.
George Kong is a Professional Services partner responsible for delivering Teradata Aster services.