Problems which were easily described to humans could not be easily handled by machines - that is to say, you could spend 10 minutes explaining a need to a programmer, and it would take a month for the programmer to design, write, test, and debug a computer program which satisfied your need.

In 1991 Grumbine began addressing these problems by building software that would operate at a higher level of abstraction than conventional software did. This software largely removed the programmer from the cycle by allowing the need to be described directly to the computer, rather than to the programmer.

In 1995, Grumbine formed Innovative Computing, Inc. to develop and market software which operated in this manner. Our first product, named MAST (Multi-dimensional Analytical and Simulation Tool) began running in production in 1996. We are still using it today, although it has had many upgrades since it began. Slicing-and-dicing was chosen as the foundation for MAST, because any question about data can be answered by slicing-and-dicing.

MAST has historically been used in the telephone industry, to analyze the customer base, to simulate new pricing models based on historical calling patterns, and to generate customer lists based on their calling behavior and other characteristics. In order to demonstrate MAST to the world outside of the telecom industry, and also to prove that MAST has capabilities well beyond the telephone realm, MAST is now being used to analyze US Census Long Form data.

To put the value of this unique approach into perspective, we suggest this exercise: First ask your IT staff to slice-and-dice dollars, costs, and transaction counts by 6 different attributes (e.g. store location, product color, weight, price range, size of order, etc.). Your IT staff will probably take several weeks to write a special program. Then ask us to slice-and-dice incomes, person counts, household counts, etc. by 6 different items from the long form census data. We'll probably give it to you the next day.