What is Presto Analytics?

Business intelligence daily seems to encompass more data from very different sources: social media for your customers, increasing volumes of real-time data for monitoring operations or managing risk, plus all the traditional historical and transactional data of your organization. Managing this 'fire hose' of data and making sense of it requires:

Many, flexible ways to visualize insights so they tell a clear, evocative story.

Presto Analytics leverages many of the existing features in Presto to do this and adds features to let you solve this problem:

Presto Analytics In-Memory Stores: to handle 'big data' with good performance using streaming access and Terracotta BigMemory. By default, Presto includes limited use of both local heap and off-heap memory through BigMemory for the Presto Analytics In-Memory Stores.

Adding BigMemory licenses provides straight-forward, highly extensible, fast in-memory data management in either a separate server and host or in an array of distributed servers using an easily-extended architecture.

Streaming large datasets also removes other performance bottlenecks, such as keeping the entire dataset in Presto memory with a DOM to support XPath queries. Data is streamed in chunks (partitions) instead.

The Real-Time Analytics Query Language(RAQL) and Analytics Engine: to query large datasets in a simple but powerful way, without the overhead of a DOM, and easily apply analytic functions. Datasets use a flat, table-like structure to support queries.

RAQL (pronounced

) is a SQL-like query language that provides performant access and handles streaming datasets. It includes a set of built-in analytics functions and a straight-forward way to provide your own user-defined functions to meet new or unique analysis needs.

You use RAQL queries within mashups, thus leveraging the power and flexibility of mashups. Presto Analytics has extended EMML, the language for mashups, to provide streaming access to common sources for datasets as well as store datasets in or load dataset from In-Memory Stores.

Mashup Tools and the RAQLExplorer: use the RAQL Explorer to easily explore the features of RAQL syntax and the datasets you need to work with. You can also create RAQL queries for mashups in Wires, using simple graphic modelling and drag and drop blocks. Or use the full power of EMML + RAQL in the Mashup Editor.

Once you have the mashup and RAQL queries you need to work with a dataset, simply add views and create apps or workspace apps using Presto’s point-and-click wizards. Use the built-in Presto views or developers can create pluggable views and add them to Presto to meet your specific visuallization needs.

Presto Analytics and RAQL give you access to datasets from a variety of sources, including:

Datasets that Presto or other systems have already loaded in Presto Analytics In-Memory Stores

Users work with the final visual analytics that are published as apps or workspace apps. These apps are based on the mashups with RAQL queries and the views you choose to include in the app.

When users find and use these apps, the associated mashup is processed in Presto Server using the EMML Engine, the Analytics Engine and any analytics functions defined in RAQL queries. They may work with datasets in Presto Analytics In-Memory Stores or directly access datasets from any supported data source.