Experience All Your Data with the Pentaho platform.

WHY?

Distributed, diverse and dynamic data across multiple silos are some of the most challenging elements to running an effective enterprise because they wall-off the ability to contextualize upstream and downstream information and the implications to adjacent operations, which creates a more reactive versus proactive environment and inhibits the convergence and unlocking of value from data. Compounded with the difficulty of integrating different systems, including diverse data formats, protocol conversions, and edge data management, many companies are finding that streamlining this integration of all their datasources provides a huge operational competitive advantage over other companies in the same industry. Streaming data is particularly challenging because of the velocity and format of continuous data generated by an array of sources and devices.

HOW?

Just like the raw material intake into a manufacturing plant, the data integration and access function ensures that your data feed is always available and will not impede decision-making at any level of your organization. This includes:

Broad connectivity to a variety of diverse data, including all popular structured, unstructured and semi-structured data sources like:

• Operational Databases: Oracle, IBM DB2 , MySQL, Microsoft SQL Server, Postgres, IBM MQ.
• Analytic databases: Redshift, Snowflake, Vertica, Greenplum, Teradata, SAP HANA, Amazon Redshift, Google Big Query
• NoSQL databases and object stores: MongoDB, Cassandra, HBase, Hitachi Content Platform, AWS S3, Google Cloud Storage, Microsoft AzureADLS.
• Spark and Hadoop: Cloudera, Hortonworks, Amazon EMR,Microsoft Azure HDInsights, and Elastic Search.
• Business applications: SAP, Salesforce, Google Analytics.
• Files: XML, JSON, Microsoft Excel, CSV, txt, Avro, Parquet,ORC, EBCDIC (mainframe), unstructured files with metadata, including audio, video and visual files.
• Operations Technology (OT) data     integration by connecting to legacy control systems, historian     databases and production management systems.
• Accessing new and expanded data sets from IoT sensors, relational     databases, video, picture, audio, documents and other core systems.
• Managing data between hybrid and multicloud environments.

WHAT?

Just like there are machines that process raw material coming into a plant, Hitachi has software that integrates and stores data from all your IT and OT sources. This includes:

Pentaho Data Integration & Analytics delivers analytics-ready data with broad connectivity to virtually any data source or application, a drag-and-drop interface to create data pipelines and templates that execute edge to cloud. Pentaho Data Integration and Analytics software connects structured, unstructured and semi-structured data sources. This dramatically reduces manual operations, shortens time to delivery and increases the performance of data extraction, load and delivery processes. With this tool, data is available and ready for consumption by business and analytics users, as well as applications or services.

Pentaho Edge Data Integration software that ingests OT data from industrial devices and sensors. It provides data processing, reading payloads and redirecting portions to select destinations.

‍

WHY?

Data scientists spend 60% to 80% of their time on data preparation, an activity two-thirds of them dislike. These individuals are also highly paid and should be spending more time on designing machine learning models and fine tuning them for use in the production environment. This work is tedious with many repetitive steps that can be automated out of the workflow, reducing the time-to-value for developing ArtificialIntelligence (AI) and Machine Learning (ML) applications.

HOW?

In a factory, as parts and assemblies move down the production line, they become more complete and valuable as finished products. A Data Operations (DataOps) platform works in a similar way by preparing data for business users including process tag-name rationalization, time stamp alignment and data range validation, allowing for example, quality and maintenance departments to work with data correlated to product or asset information as defined by the organization. Data activities that can be automated include: 

• Data contextualization and normalization performed close to the office, facility, machine asset or edge.
• Scheduling of data processes. These activities determine the data transfer schedule including execution, load and transform activities, making optimal use of local resources.
• Dispatching data analysis orders. Depending on the type of data needed, this may include further distribution of data queries and work orders requested by operations teams and adjusted for analysis or ML training     needs.

WHAT?

Pentaho Data Integration& Analytics centralizes the ingestion, blending, cleansing and preparing of diverse data sets from any source, in any environment with an easy-to-use, drag-and-drop data pipeline workflow designer — without code. It streamlines and speeds up data delivery to enable data self-service to collaboratively build, deploy, and monitor data flows for faster, more confident business decisions. It allows your data engineers and data consumers to collaborate more effectively and customize their own dashboard views. It blends data across lakes, warehouses, and devices, while orchestrating data flows in hybrid and multi-cloud environments with improved dataflow visibility and tools to customize your data.

WHY?

Over 90% of business data is dark and unusable for broader purposes. In many cases, it is dirty, untrusted, not standardized and lacks context. Data is also inconsistent and has little context to explain what it is, where it came from. This further adds to a lower reliability score. It cannot be used in its raw form for critical operations decision making. Many times, the process to organize and enhance the data for business use is highly manual and error prone as well as taking up valuable time from workers who could be focusing on more meaningful work for the organization.

HOW?

Just like a manufacturing organization receives raw material through the incoming department of the plant that must be graded, curated, categorized and stored in the appropriate bins to be utilized in the production process, data also needs to be identified, classified and made available fora specific use and purpose. This includes: 

• Using AI-driven discovery and profiling through unique data fingerprinting to automate discovery and classification of structured, semi and unstructured data.
• Discovering, identifying and classifying unknown or incorrectly labeled metadata that can result from poor data lineage, data drift, or simply new data sources.
• Management of data definitions including storage, version control and exchange with other master data.
• Business users/analysts connect to data sources and can explore them visually.
• Using AI-driven discovery through unique data fingerprinting to automate discovery and classification of structured, semi and unstructured data.
• A business glossary to build or import a taxonomy of business terms and establish relationships.
• Business rules to validate data and assess conformity to business policies.
• Data quality to quickly assess your critical quality metrics across the operation.

WHAT

Pentaho Data Catalog software delivers trusted, fit-for-purpose data through automated data discovery and classification. An intuitive user interface allows you to search data, much like an on-line shopping web site, providing at-a-glance views tailored to user access and data use rights. Hitachi’s approach to machine learning finger-printing technology gives you a customizable business glossary to label data fit for all your organization’s needs. It has an award winning, out-of-the-box visual metadata report development environment allowing you to get started quickly, identifying data quality issues at the source.

WHY?

The quality of data available for analytical innovation is many times unknown and poor leading to lack of trust and reliability with business stakeholders. Companies are also finding that data privacy and regulatory compliance is becoming more important. Information that includes patented production processes and product designs, confidential employee information, emissions compliance data and customer records all need to be safeguarded. Also, the explosion of collected data can’t all be cost effectively stored in the cloud and some data is transient where it may only be needed for a few hours before being summarized and placed in long-term storage. 

HOW?

In a factory, a Manufacturing Execution System (MES) or automation system keeps track of all the parts and production processes that entail running the factory. DataOps software does the same for data trust. Trust in data requires the confidence and knowledge that your organization’s data is fit for purpose and ready to act on. Data trust is about expectation and reliability. Trusting your data means knowing your data. This includes:

• Management of data resources. This may include registration, exchange, and analysis of data resource information, aiming to prepare and execute data operations. quick assessments of critical quality metrics ​across the business.
• Data Lineage: Surfaces hidden lineage from your data, providing users with additional insights to select the best available data for their projects.
• Fully discover sensitive data with exceptionally low manual effort
• Explore and discover hidden data relationships and characteristics easily and intuitively, all in the language of your business.

WHAT?

Pentaho Data Catalog intelligently automates profiling and discovery across your data estate of applications and data stores, edge, on-prem and multi-cloud. Make discovered data actionable by leveragingAI-driven capabilities to give business context to your data and apply tailored business rules that ensure adherence to your organization’s privacy, compliance and security requirements. Harness a rich metadata library exposing hidden relationships, finding data in an economical and intuitive ways, assessing data quality, improving reliability and informing fit-for-purpose. This allows organizations to engineer reliability and adhere with their governance, security and compliance controls quickly and accurately before data can be used, thus reducing the risk of exposure.

WHY?

Ninety-two percent of organizations plan to deploy predictive and prescriptive analytics more broadly, while 50% have difficulty integrating them into existing infrastructure. To facilitate the human in the loop for analytics, an easy ability to visualize information on dashboards and to receive alerts needs to be in place to experience value.Data science resources are also at a premium so the ability to quickly assemble machine learning datasets frees up their time to focus on the outcome of the analytic rather than data assembly, which speeds time to value. In fact, most enterprises struggle to put models to work because data professionals often operate in silos and create bottlenecks in the data preparation and model update workflow. 

HOW?

The ultimate step in the manufacturing process is packaging and labeling, which makes the product useful and consumable by the end user. The same is true of data analysis and visualization of information where a HumanMachine Interface (HMI) or dashboard can be created quickly to meet a specific job role or departmental needs. This includes: 

• Embedding analytic assets into other web applications.
• Ability to train accurate and dependable ML equipment models.
• Aligned with existing back-end requirements: security integration, single sign on, multi-tenant deployment.
• Analytics dashboards for performance analysis. Displaying useful information out of the raw collected data using trained models to provide predictive and prescriptive analytics.

WHAT?

Pentaho Business Analytics provides a spectrum of analytics for all user roles, from visual data analysis for business operations to tailored dashboards for all audiences —from business executives to front line workers. Users can create reports and dashboards as well as visualize and analyze data across multiple dimensions without dependence on IT or developers. The Pentaho platform streamlines your entire machine learning workflow and enables teams of data scientists, engineers and analysts to train, tune, test and deploy predictive models.