Analyst Perspectives

Teradata introduced some enhancements to its Vantage platform last year in which they expanded its analytics functions and language support, and strengthened tools to improve collaboration between data scientists, business analysts, data engineers and business personnel. Some of the key enhancements included expanding the native support for R and Python, extending the ability to execute a wide range of open-source analytics algorithms, and automatic generation of SQL from R and Python code. These updates are included to reduce data silos, enabling a wide range of data and analytics personas to collaboratively run complex analytics in a self-service manner.

Teradata is a Relational Database Management System (RDBMS) for developing large-scale data-warehousing applications deployed on premises or in the cloud. It offers rich workload-management capabilities that can support multiple data-warehouse operations at the same time and can handle large volumes of data and can be scaled up to a maximum of 2048 nodes. Teradata’s architecture is based on Massively Parallel Processing (MPP), which divides large volumes of data into smaller processes and executes them in parallel.

Emerging use cases of advanced analytics and ever-increasing types and sources of data are compelling data scientists to utilize varied data-science techniques. These techniques include multiple user languages such as SQL, R, Python, and Scala, as well as various analytics technologies such as machine learning (ML), graph analytics and neural net analytics. This is combined with multiple styles of analytics that span statistical, text, ML and deep learning. The need for operationalization at the speed of business adds another layer of complexity, which can be critical for data-powered and analytics-driven decision making. We assert that through 2024, relational data warehouse technologies and big data platforms will converge to create enterprise data platforms, enabling organizations to collect and analyze all types of operations-generated information.

Teradata Vantage is an example of an enterprise data platform. It combines multiple data-storage engines with multiple analytics engines to enable both SQL and more advanced analytical processing over relational data and big data storages. Teradata Vantage delivers analytics functions and engines, tools and languages, and supports multiple data types. It embeds analytics engines close to the data, which eliminates the need to move data and allows personnel to run analytics and models against larger data sets without sampling.

Teradata Vantage’s Machine Learning Engine (ML Engine) provides a suite of ML functions to perform analytics on all available datasets. It can support use cases that require training the model with a full dataset, without giving up end-to-end performance of an analytics workload. Teradata’s ML Engine uses SQL-MapReduce Collaborative Planning, which was part of its Aster Data acquisition, to enable multiple types of analytics. It can work with a combination of structured, multi-structured and unstructured data (such as text, numerical, tabular, files, CLOBs, and BLOBs data types) to perform path and pattern, ML, and graph analytics in the same workload.

Organizations are collecting and analyzing data to forecast future events in order to improve various aspects of their businesses. The process requires the fluidity to manage all types of data sources and move between different analytics, exchanging data and gaining insights as the data is generated. Teradata is an open and scalable database management system that offers linear scalability, which allows a large volume of data to be handled efficiently at one time by adding nodes for increased data workloads. Business personnel can connect to Teradata with various business intelligence (BI) and analytical tools for high-performance analyses at scale. I recommend that organizations looking to integrate data from various sources and simplify their data and analytics ecosystem should consider Teradata when evaluating vendors.

Regards,

David Menninger