Xcalar Data Platform accelerates your analytics cycle from the first view of a data source, through prepping the data and refining data quality, to the operational use of the data across your organization.
Both data prep and data quality result from iterative analytics cycles. The analytics cycle for data prep consists of the following steps:
Profiling data fields
Cleansing data fields
Transforming data fields for use
Combining disparate fields and data sets
In addition, enterprise data quality results from the following secondary cycle:
Applying your business rules to data at scale
Surfacing data anomalies
Using data lineage to diagnose anomaly sources
Refining business rules to resolve anomalies
Typically, analytics departments use a suite of tools for analytics. This results in slower analytics cycles, due to unnecessary data copies, inefficient processing, and broken data lineage.
Xcalar Data Platform is a hybrid cloud and on-prem solution that provides users with visual programming tools to rapidly build a data model. Creating a model results in a dataflow, which provides the following:
The data’s lineage, which is the necessary information to follow schema evolution and trace data anomalies back to the source
The algorithm to apply the data model efficiently at scale to meet SLAs
Xcalar Data Platform makes developers more productive by focusing their use of custom code for new data formats, proprietary logic, and machine learning algorithms. Through the use of visual modeling tools and focused custom code, Xcalar Data Platform increases the velocity of each analytics cycle, resulting in more iterations in less time and, therefore, higher data quality.
Xcalar Data Platform works directly with source data files using metadata, without copying data into an internal format.
Xcalar Data Platform works with structured, semi-structured, or unstructured source data of any format, from file or streaming sources.
Xcalar Data Platform provides efficient data profiling on terabyte scale data with statistical summaries, histograms, and data correlations.
Users interactively create data models using a spreadsheet-like user interface; resulting models track lineage of data from sources through its transformation.
Users apply business logic on data in Xcalar Data Platform by writing and applying a list of rules to surface anticipated data anomalies.
While processing each data operation, Xcalar Data Platform surfaces the anomalies inherent in the data; users can triage these at any stage of analytics work.
Xcalar Data Platform’s responsive interface performs interactive analysis using relational operators on up to 100 billion rows.
Data scientists can train and deploy ML or predictive algorithms across petabytes of data at any stage of the data pipeline.
The algorithm resulting from modeling is displayed as an auditable graphical diagram of operations; it can be saved, operationalized, and scheduled to run on production data.
Xcalar Data Platform processes read and write operations with near linear scalability while maintaining strong data consistency, across cloud-scale clusters.
Analysts pull data via optimized JDBC queries using BI applications, such as Tableau, Qlik, and Power BI for visualization of data.
Visual programming provides well-articulated opportunities to apply proprietary logic to data models, using Python.
Large scale analytics workloads are run in high-throughput mode to meet performance goals. Xcalar Data Platform allows dynamic skew detection and dynamic WL management.
Xcalar Data Platform supports integration with Kerberos, LDAP, OAUTH, and custom authentication services for authentication and user management.
Xcalar Data Platform users can easily share workbooks, datasets, and custom code to jointly solve problems.
All operations, including user code, are run in separate containers for robustness and stability; restart workloads using automated system recovery logs.
Refine your data and accelerate your analytics cycle.Learn more
Meet your SLAs when processing micro-batch updates.Learn more
Unlock your data lake for analysts and data scientists.Learn more
Access all your data without data movement.