Auguri Server

AuguriKernel

Auguri provides a decision support platform (DSS) that offers an environment for the development, deployment and implementation of web based applications that require the services performed by the AUGURI Server. At the heart of Auguri’s platform is the AUGURI Server. A data server that performs a series of web-based services including tradeoff based searches, inference, expert knowledge capture, profile generation, analytics and reporting services. The AUGURI Server is the kernel of all solutions.
Auguri’s technology powers a slew of new functionalities to traditional applications including Procurement, SCM, HRMS, CRM, ERP, B.I., ISS etc….

The AUGURI SERVER is the brain of the Auguri product, its service oriented architecture and support for open standards makes it robust, scalable and reliable. It is designed to be data source independent. It provides a slew of web-based functionality for intelligent data access and analysis. This functionality includes the ability to perform inference, tradeoff-based search, expert knowledge capture, profile generation, analytics and generic reporting services. The Auguri server operates using advanced scoring mechanisms, rather than straightforward SQL implementations, thus delivering a high degree of relevance and control across the entire spectrum of functionality.

AUGURI’S ENGINE

A decision (selection, search etc) is the selection of the optimal alternative(s) from a set of choices. The decision is typically made by performing tradeoffs (weighing the relative importance) over a set of factors or criteria. Tradeoffs become necessary in situations where various factors are conflicting, and alternatives are consequently less than perfect. To give a simple example, a consultant that often travels to deliver presentations is interested in a light weight laptop with a large screen. However, the criteria are conflicting; the larger the screen the heavier the laptop. Hence the ability to tradeoff between conflicting criteria is paramount.

The objective of a decision, a selection, a prioritization, or a triage is to select the best solution(s) from a set of alternatives. To be able to make a selection, the first component required is a set of data, or alternatives. This is the database of options which will be ranked according to their weighted proximity to a hypothetical ideal result.
A decision or selection is typically based on a set of criteria. These criteria are encapsulated by their criteria behavior (shown on the left side of Figure 1) and embodied by a function, such as a “utility” function, that captures the way we think about that criterion. For example, screen-size is a criterion when selecting a laptop, and typically, its behavior is captured by the notion that “larger is better”. Weight is another selection criterion, defined by a different function since we typically seek lighter laptops. But because laptop weight typically increases with screen-size, we have a conflict between these criteria. Humans know only one way to mitigate these conflicts: Tradeoffs.

The ideal laptop is lightest one that has the biggest screen. Unfortunately, such a laptop is hypothetical and does not exist. Criteria are almost always conflicting in real-world scenarios. This conflict is addressed via the notion of tradeoffs – relative prioritizations. Tradeoffs are a central component in the Auguri platform, just as they are in real decision-making situations. They play a pivotal role in determining how various options will score.

AUGURI’S INFERENCE ENGINE

Implementing BI solutions or rule-based engines to analyze alternatives in light of a user’s preferences can be resource intensive both initially and during ongoing maintenance. Statistics-based engines may not contain enough detailed information to enable calculations at the level of individual user preference. More problematic is the reality that SQL engines are unable to infer the issuing query based on search results which means pragmatically an inability to understand the rationale or reasoning behind particular results.

The Auguri Inference Engine can reverse engineer your queries. It takes a ranking of a set of possible alternatives (or result set) and determines the relative importance of various criteria accordingly (or the original query). Based on the same technology as WISE, the Auguri Inference Engine can not only tell you what criteria are important, but how important each criterion is on an individual basis. Inferring query parameters from the results of a query provides a particularly useful tool for analyzing application and user behavior and preferences. The inference capability opens up the door to a whole new suite of applications. For example, by viewing a user’s click-stream, one can infer the not only the nature of the visit but also the motivations behind it. Auguri’s state-of-the-art inference technology provides applications, in addition to what data is searched, intelligence about why it is sought.