CeleriQ Overview

The CeleriQ indexing engine is a framework to serve up structured information quickly. In contrast to a relational database, which is not structured for easy searching, CeleriQ allows users to find relationships in information without a cumbersome SQL language. Data can be filtered, searched, and aggregated by using our Piloted Search technology.

Relational database are very slow when combining table data. This has the effect of making queries on large sets of data very slow. For a company website or other mission critical, customer-facing applications, a need exists for very fast retrieval of large sets of data. An example of this might be a database of all used cars in a state or even the whole country. The database might have hundreds of thousands or even millions of rows of automobile information. If you built a customer-facing website to display this data, a user expects to search for autos and retrieve results almost instantly. He will not wait for 30 seconds while the application joins tens of tables to find the appropriate information. He will most likely just go to another website; a faster one.

To fill in the performance gap, CeleriQ provides an object oriented way to query your data quickly and get the results on the screen almost immediately. The secret is to take your data and index it in a new way with the CeleriQ engine. The engine is compatible with SQL Server and works with your existing SQL Server databases.

CeleriQ allows you to specify the data needed in a timely fashion and define how the data will be used. After the initial data setup, the indexing engine will build indexing tables in a SQL Server database. This organizes information differently than a relational database and improves performance. Another benefit is that software developers will find accessing information is very easy. There is an object oriented API that returns records, paged and ordered, ready for display.

In conjunction with Piloted Search, the Forward-Knowledge capability allows applications to display the number of listings down all paths. When many items exist in a category of information, such as state, it is useful to know how many items will be returned when filtered. For example, when listing US states, there may be about 50 items listed but in normal applications there is no way to know how many listings are in each drill-down. The engine will return not only the state text information but also how many listings are in each one. Applications can now inform users of the listing count before actually navigating down that path.