Pentaho update aids data integration, semantic modeling

Speed, simplification and security

Many enterprises have made AI the focus of their application development initiatives since OpenAI's November 2022 launch of ChatGPT marked significant improvement in generative AI technology. AI applications, however, require far more data than traditional analytics reports and dashboards to be accurate.

This is an incremental improvement in some critical areas, most notably ease of use, governance and performance. Market demands are rising in all three areas as enterprises adopt AI to democratize data consumption and streamline or enhance business processes without incurring significant risk.

As a result, the volume and complexity of data workloads is increasing.

The capabilities that comprise Pentaho's platform update are designed to help users better manage larger and more elaborate data workloads that provide a foundation for AI development, and were prioritized based on customer feedback, according to Sandeep Prakash, the vendor's vice president of product management.

"Version 11 has a good balance of features based on customer requests and elements we know customers will benefit from as they manage heavier data workloads," he said.

For example, the new user interface was developed in response to user feedback while the browser-based version of Pipeline Designer is a feature aimed at easing burdens on data engineering teams, Prakash continued.

Pipeline Designer is part of Pentaho Data Integration and is a feature that enables users to create pipelines for tasks such as extract, load and transform (ELT) workflows. The browser-based version simplifies pipeline development by removing local installation requirements -- configurations that need to be set up on local systems -- and includes a new interface for creating jobs to make it more accessible to business users.

Project Profile likewise addresses pipeline development. But rather than simplify individual data integration jobs, it enables Pentaho users to group related jobs, transformations and configuration files into containers to reduce deployment complexity and better enable collaboration.

While Pipeline Designer and Project Profile simplify building and managing pipelines, Semantic Model Editor is aimed at making it easier to model data. The new tool replaces Schema Workbench and Data Source Wizard with a modernized means of creating and managing semantic models that standardize defining data's characteristics -- its metadata -- to make it easier to discover and operationalize relevant data for a given initiative.

Meanwhile, new authentication that integrates with identity providers such as Azure, Google and Okta, and redesigned permission controls both address governance and security.

Perhaps the browser-based Pipeline Designer and Semantic Model Editor are the highlight features of Pentaho's platform update given that each simplify complex processes, according to Catanzano.

"Pipeline Designer removes the need for local installations and offers a streamlined, intuitive interface, making it easier for distributed teams to collaborate and accelerate pipeline development," he said. "The Semantic Model Editor modernizes the analytics experience by replacing older tools with a cleaner, web-based workflow, ensuring a smoother transition for users while enhancing usability and governance."

Petrie, meanwhile, highlighted Project Profile because it helps enterprises standardize data consumption across environments such as multiple clouds and on-premises systems.

"It gives data and DevOps engineers modular, containerized pipelines that they can reuse on various platforms to speed data readiness and reduce governance risk," he said. "This helps simplify data consumption across hybrid and multi-cloud environments, which is to say most data environments."

Looking ahead

With Version 11 of its Data Integration and Business Analytics platform update now available, Pentaho's product development roadmap is focused on helping customers build a trusted data foundation for AI initiatives and providing customers with automation and natural language processing capabilities to improve productivity, according to Prakash.

"We see our roadmap aligning with customer needs that fall into two categories -- data for AI, and AI for data," he said. "Over the coming quarters, you'll see us deliver capabilities around AI-enabled discovery, semantic search for data [and] building … agentic workflows."

Petrie noted that Pentaho's data integration and analytics capabilities are generally in line with those of its competitors. However, one way the vendor differentiates itself is with data optimization capabilities that help customers identify and archive less-used datasets to reduce costs. Creating messaging that emphasizes Pentaho's unique capabilities would be wise, Petrie advised.

"I'd be interested to see Pentaho play this up more in their sales and marketing efforts," he said.

Catanzano, meanwhile, suggested that Pentaho could continue serving its current users and perhaps attract new ones by adding features and integrations that better enable customers to develop AI tools.

"To continue evolving, Pentaho could expand its AI and machine learning capabilities by integrating with popular AI frameworks or offering pre-built, industry-specific AI models," he said. "This would not only enhance its value for existing users but also attract new customers seeking to accelerate their AI adoption."

Eric Avidon is a senior news writer for Informa TechTarget and a journalist with more than three decades of experience. He covers analytics and data management.