Direct is the Fastest Path to Insight
The Incorta Direct Data Platform™ gives any business user the ability to analyze complex, full-fidelity business data in real-time. It provides integrated user access to all components of the data pipeline — data connections, business semantics, security settings, scheduling and publishing — allowing lines-of-business to be more self-sufficient and more agile.
Incorta customers skip past costly and time-consuming data warehouse projects, and focus on delivering data, insights, and real business results.
Direct Data Acquisition
As Incorta loads data, it creates a dataset that is actually smaller than the source data, as opposed to most dimensional modeling schemes which create a data set that is larger than the source data. This means efficient cluster data transfers, and fast data loads into memory.
- Data Connectors — Easy to configure and connect to every database, enterprise application, data stream, and data file format. Can monitor and ingest data lakes. Extensible to any data source with our SDK.
- Parallel Data Loader — Spark executors over partitioned data sources.
- Schema loading and Introspection — Each connector can interrogate the data source for column names, data types, null handling, cardinality, and relationships.
- Physical schema — A data catalog of physical structures that determines what data will be loaded from a data source. Schemas also include data relationships, filters and computed columns.
- Direct Data Mapping — preprocesses raw data to determine all potential query paths, enabling lighting fast queries on normalized (application) data models (contrast with dimensional, aggregated data models, cubes) meaning no data reshaping or transformations required before running analytics.
Direct Data Analytics
While most other solutions just compress data on-disk and uncompress it as it is read into memory, Incorta maintains a columnar and compressed data organization both on-disk, and in-memory. This makes it possible to analyze 10x to 20x more data than would normally fit into available memory. However, Incorta is not memory-bound. Data is dynamically loaded from disc as needed, unless a data table has been specifically configured to always remain in memory.
- Memory Optimized Analytics Engine -- In memory, on disk, remote/virtual; Data is brought into memory compressed meaning 10-20x memory efficiency.
- Breakthrough performance on raw, normalized data via Direct Data Mapping
- Data Exploration and Publishing
- Data Analyzer is a drag-and-drop, easy-to-learn data discovery and visualization tool that leverages the business semantics and boasts a wide variety of visualizations.
- Dashboards group visualizations, data tables and KPIs for data consumers. Users can personalize the data shown, apply filters and can click to drill into details.
- Native iOS application for mobile users.
- Data Catalog -- Incorta Business Schemas are aggregations of one or more physical schemas to represent business concepts and simplify the data model for business users. Can be created without code using the data Analyzer.
- SQL access into both the in-memory analytics engine and the on-disk data hub for 3rd party BI and analytics tools such as Tableau, PowerBI and others.
- Security -- User accounts, groups and roles; Role-based access controls; Encryption; Attribute-based dynamic row filtering; Single Sign-on (SSO); Auditing
“Smart” Data Lake
- Data Lake architecture but with metadata, upsert capability, multi-version read consistency and
- Data Lineage -- Ability to discover source of computed or compound metrics and perform ‘where used’ analysis. Report on changed or missing objects.
- Open standards -- No vendor lock-in (Parquet, Spark, SQL, PySpark, Scala, R, Zeppelin)
- Shared Storage -- Data and metadata is stored on elastic cloud services such as Amazon S3, Microsoft Azure Data Lake Services, and Google Storage.
- Single source of truth — engineers, data analysts, data scientists, and business users all access and work against the same data