No Decisions Until It's Time

Posted by Sam Chance on Apr 27, 2017 7:53:00 AM
Find me on:

Superior decision making is an essential aspect of life, whether it be in business, national security, health, environment, and every other aspect of human existence. Look no further than geopolitical affairs, such as North Korean or Iranian relations, to understand the importance and impact of decisions on human life, indeed the entire planet. This is not an exaggeration. From an Information Technology point of view, the goal is to provide complete and accurate information on demand to support decision making.

soft-2032762_640.jpgContemporary decision making strategy stems from the “just in time” philosophy adopted in logistics and supply chain management. We’ve seen this approach captured, for example, in the phrase “No Decision until It’s Time” (NDIT). The idea is to defer the resulting course of action(s) until exactly the right time to keep options available as long as possible. In other words, we want to avoid early commitments. Having options means we enjoy flexibility, and can adapt based on the latest information. Adaptability and flexibility are strategic goals for almost any organization – commercial, military, government, and others. Consequently, Information Technology must be adaptable and flexible as well.

Consistent with NDIT, Anzo Smart Data Lake® (ASDL) seeks to minimize early commitment to help maximize adaptability and flexibility. One method of avoiding early commitment is ASDL’s standards-based graph data model that expects change. More specifically, ASDL allows enterprises to load source data without committing to application-specific data models. ASDL elicits the source data schema, and automatically generates a graph model based on the implicit schema. Then, users may build application-specific models and custom mappings, or late-binding, based on their latest needs. Additionally, avoiding custom mappings on ingest removes the time-consuming and brittle practice of creating mappings of source data on ingest, where it is least valuable.

In addition to avoiding early commitment to data models, ASDL allows end users to create additional data transformations in the “last step” before data is rendered in user interfaces. Users enjoy the option to create calculated values, and even store them as computed properties, which other authorized users can then apply.

Another way ASDL supports NDIT is “emergent questioning” wherein queries are not pre-defined for application-specific requirements. In other words, ASDL allows analysts to ask previously unknown questions when they discover the need. Our graph model is built for flexibility and context formation, which are essential to efficient and effective decisions, respectively. Flexibility allows for rapid change, which in turn promotes quick response times. Context formation allows ASDL to answer questions more completely, which means better decisions.

In addition to avoiding early commitments in models and questions, ASDL avoids unnecessary commitments to computing resources through elasticity. A given deployment of ASDL may expand and contract computing resources based on demand, budget, priority and other factors. For example, a given deployment of ASDL may require greater query capacity during normal business hours, and more ingestion capacity during night hours. Outside these two periods it may make sense to retire certain resources to save resources. Like information on demand, ASDL provisions (and retires) computing resources on demand, thus saving money and meeting demand.

In summary, Cambridge Semantics designed and built Anzo Smart Data Lake to support decision makers who make critical decisions, but not until it’s time. ASDL empowers adopters to avoid early commitments to data models, questions, and computational resources, which, when combined, provide the right information at the right time.

To learn more about Cambridge Semantics and the Anzo Smart Data Lake, download the Bloor Group's vendor profile here.

Download the Report

Tags: Smart Data, Semantics, Big Data, Data Lake, Graph, Smart Data Lake