Market leading business intelligence and data visualization tools are not designed to handle the large data volumes typically associated with “big data”. At the same time, data sources like Hive and NoSQL are not able to deliver sub-second response times for complex queries.
Users can visually design star/snowflake schemas and build high-performance OLAP cubes. Data analysts can drag and drop facts, dimensions, and measures. Infoworks automatically builds a fully pre-aggregated and optimized cube natively on the big data platform, providing sub-second response times to most user queries.
Big Data queries are dramatically accelerated by Infoworks through query interfaces that support a variety of analytics use-cases across business intelligence, data science, ad-hoc and batch. For BI analytics, reporting, and dashboard style use-cases, Infoworks cubes provide sub-second and interactive response times. For ad-hoc queries, the in-memory accelerated models provide fast access to granular data for a variety of use-cases, while batch use cases can benefit from the optimized data models. Infoworks provides these multiple layers of query acceleration to deliver the right performance and scalability characteristics for each use-case.
Infoworks automatically propagates upstream changes down to in-memory and data lake data models. Whether the change is the addition of a new source column, change data capture updates of content, or modification of transformation logic, changes can be automatically propagated to the downstream data models.
Infoworks in-memory models provide much better query performance when compared with Hive. The following benchmarks use the TPC-DS dataset run in memory. All queries were run without modification resulting in performance that ranged from 150x to 500x faster than Hive.