Decisions & BI



Traditional Analytics tools often fall short of delivering on their promises. They can be expensive in both time and resources, and fail to achieve the efficiency and adaptability which are essential for organization to be effective in a competitive market landscape.

Typically, these tools will exist in one of two flavors: The first is a data-centric approach, which may utilize query modeling, statistical analysis, ETL, pivotal transformations, dimensional cubes, and dashboards. The idea is to provide a “big-picture” person with a high-level perspective of the entire organization, a perspective compiled from a vast array of data collected over time. While this approach is effective in summarizing large data sets at a high level of abstraction, it provides little guidance or insight in terms of the factors influencing the data. This component is left up to high-level personnel, who may be forced to make decisions without adequate or appropriate levels of detail.

The second approach is programmatic, rather than data-oriented. It is generally reliant on experts who devise specialized programs to handle specific problems. Techniques here may involve artificial intelligence, neural networks, rule-based systems and decision trees. While this approach is effective in some scenarios, it suffers from being resource-consuming and somewhat inflexible. Even a small change in an organization’s priorities or structure can demand a complete reworking of a programmatic solution.


Auguri bridges the gap between these approaches by analyzing data within the context of expert-defined situational variables. The end result is an Analytics engine which is “decision” oriented rather than data-oriented. More importantly, the Auguri solution operates effectively in real-time, and has the flexibility to adapt quickly to new data and new scenarios.

One key advantage of the Auguri solution is knowledge capture. With traditional solutions, expert understanding of the many criteria behavior functions is lost when employees leave the organization. With Auguri, this knowledge is automatically captured and embedded in the engine. Consequently, Auguri is superior in its ability to facilitate knowledge sharing, which can take place across an entire organization. This raises everyone involved to the level of the expert. The end result is a highly collaborative engine that combines the expertise of many individuals at many levels.

The second advantage of the Auguri solution is its ability to provide real-time tradeoffs and “what-if” analysis. In particular, the Auguri inference engine allows users to quickly analyze how various decisions are affected by decision criteria.

By providing an efficient, real-time analytics engine which facilitates knowledge capture and sharing across an organization, the Auguri solution is adaptable to a wide variety of scenarios.

Business Intelligence

The objective of business intelligence is to provide a supporting infrastructure to the operations of an organization. This translates into allowing all individuals to make the best decisions in their day to day activities. This means selecting the optimal solution or course of action out of a set of possible alternatives. Decisions are usually based on a set of criteria. These criteria are often conflicting, i.e. we want a lightweight laptop with large screen, but the larger the screen the heavier the laptop. As a result decisions are often based on tradeoffs and compromises.

In addition, often decision makers have to deal with uncertainty. A corporate strategist does not know with certainty the impact a price cut will have on the market or the response of its competitors.  The data used in the decision process is usually incomplete and marred with inaccuracies. This all makes the decision pretty complex.

To make the right decision, it would be great if you could have at your fingertips the know-how of the experts. And that knowledge is not always handy.

Challenge: Traditional Decision Support Systems and Business Intelligence tools fall short on delivering on their promises. Their key challenges stem from the fact that they are focused on retrieving data along certain dimensions, summarizing and enforcing certain rules to assist in the decision process. These antiquated techniques present the following problems:

  • They are expensive. The high cost of traditional techniques stems from the need to continually program the system to assess new alternatives.
  • They are time consuming. They require either pivotal transformation, or significant programming or they rely on trial and error techniques.
  • They fall short on functionality
  • They do not handle uncertainty
  • They cannot address both complex and simple decisions
  • They lack true collaborative decision making capabilities

Adding Decision to your BI system: The purpose of any decision support system is to reduce the time it takes to make a decision as well as provide the necessary information and tools to optimize the decision. Today’s analytic tools use either a data-centric approach or a programmatic approach.  The first relies on query modeling, dashboards, and other data-displaying apparatuses that provide a detailed, high-level overview of an organization.  The benefit?  This approach effectively abstracts large sets of facts and figures in an easy-to-understand array.  However, the data gathered provides little insight into what has influenced those facts and figures.  Decisions regarding the collected data are left up to high-level personnel, who, by the very nature of the data-centric approach are left without key details that would significantly assist in the decision-making process.

What of the programmatic approach?  The programmatic approach targets specific problems through the use of specialized programs.  The use of artificial intelligence, rule-based structures, decision trees, and neural networks are implemented to fill in gaps of knowledge necessary in decision making.  The problem?  Small changes in an organization’s priorities or constitution can mean complete reprogramming.  Programmatic solutions, while effective, are time-consuming, inflexible, and overly sensitive to every little change.
Auguri bridges the gap between the traditional approaches by leveraging the intelligence embodied by the data and tapping on experts’ knowledge to focus on the decision making process and optimize it. Auguri enables a slew of new solutions that require real time decisions.

Trade-offs and inference are paramount to any decision support software, because the way we think when we make decisions is by analogy (I want a laptop like Bob’s) and by tradeoff (I prefer the blue car but will take the red one if I save $2,000).

Ask yourself these few questions.  Are you well-served by systems where the end-users require extensive training to be productive?  Are you well-served by systems where the user may be either inundated by vast quantities of irrelevant data, or conversely is starved for valid alternatives because they didn’t, or couldn’t, specify valid data ranges?  Are you well-served by systems where the end-users may not be able to easily identify the optimal solution to their question because of the complexity of the data presentation?

Shouldn’t Decision Support Systems be designed to yield quick and easy, but accurate and optimal Decisions?  Decision Support has come a long way, but the ultimate goal should be Decision Systems.  To do this, you need to align and leverage your DSS systems with the human trade-off based decision making process.

It is time to put the DECISION into your Decision Support Systems!