Industrial companies stand at a pivotal moment in their digital transformation journeys. AI promises to revolutionize how they leverage data and configure systems, but the path from proof-of-concept to practical application remains challenging. HighByte Intelligence Hub version 4.3 bridges this gap, delivering capabilities that help industrial enterprises become AI-ready while reinforcing the core functions required for reliable operations.
This release addresses three critical needs: high availability for mission-critical operations, deeper integration with industry leading enterprise cloud platforms like Microsoft Fabric and Databricks, and MCP Services for governing industrial AI adoption.
Stateful Pipelines and Enhanced Processing
High Availability
For regulated industries—life sciences, food and beverage, oil and gas, utilities—data integrity directly impacts compliance and revenue. Without reliable data collection, production output becomes unsellable or requires extensive manual certification that erodes margins.
The new High Availability (HA) mode continuously synchronizes both configuration and state across Intelligence Hub runtimes through a common PostgreSQL database. When a primary hub becomes unavailable, the secondary hub assumes control with full awareness of the latest configuration and pipeline activity, ensuring seamless failover with minimal data loss.
With state synchronization in version 4.3, organizations can now engineer stateful integrations (pipelines that track what's been processed and maintain context from previous runs) while meeting their high availability requirements, ensuring reliable operations even in the face of hardware or system failures. Learn more about this feature with HighByte CTO Aron Semle in his demonstration of HA mode on YouTube.
Advanced Data Processing
Version 4.3 of the Intelligence Hub introduces powerful new pipeline capabilities for complex integration scenarios. Two new Control stages—'For Each' and 'While'—enable looping operations. The 'For Each' stage iterates through arrayed data structures, and the 'While' stage continuously iterates until a condition is satisfied, which is especially useful for multi-step and multi-system API integrations.
Figure 1: Looping stages in HighByte Intelligence Hub version 4.3
The new JSONata stage uses simplified expression syntax for manipulation of data structures for reshaping, filtering, mapping, and aggregating datasets. The Delay stage introduces pauses within pipelines, allowing external processes to complete or catch up before proceeding to the next stage in a pipeline—a practical solution for real-world integration timing challenges.
Deeper Interoperability with Microsoft and Databricks
HighByte Intelligence Hub has become the data engineering companion for modern Data Lakehouse platforms, serving as the premier bridge between factory operations and ready-to-use data for analytics and AI. Version 4.3 deepens this integration with Microsoft Fabric and Databricks.
The new OneLake connection delivers data directly into Microsoft Fabric's unified storage layer, allowing industrial enterprises to bulk load contextualized data. Input support for Azure Blob Storage enables bidirectional data flow, retrieving schedules, recipes, or numerical control files from cloud applications for use in industrial automation systems.
The new Databricks SQL connection enables direct data retrieval from tables within Databricks' SQL warehouse, supporting "cloud-to-edge" and "reverse ETL" patterns. Organizations can now operationalize analytical and ML/AI insights from Databricks in real-time on the factory floor.
MCP Services for Industrial AI
Since HighByte released the first MCP Server for industry in version 4.2, a variety of industrial technology vendors have released their own MCP Servers. This proliferation introduces new challenges around agent best practices, governance, and security that version 4.3 addresses. The latest release also offers new opportunities to utilize AI to configure and manage deployments.
Data Tools for AI Agents
Intelligence Hub version 4.3 provides turnkey MCP data tools for OPC UA and MQTT connections. When enabled, these tools automatically expose browse and read tools, providing an out-of-box experience for experimenting with AI agents and discovering data through natural language.
The new MCP Client connection allows enterprises to consolidate and control access to third-party MCP Server solutions, governing which AI agents can access which systems while enabling innovation.
MCP Services
The new MCP Services view provides centralized management of all MCP tools, including Pipeline-based tools, connection-based tools, and external MCP Server tools. Users can view each tool's name, description, status, and origin, and navigate directly to its underlying configuration.
Figure 2: The new MCP Services view in HighByte Intelligence Hub version 4.3
Configuration Tools: Building Pipelines with AI
Version 4.3 introduces the first configuration tools for the Intelligence Hub, exposing Pipeline configuration to AI agents. Users can now prompt agents to create and update Pipelines using natural language, dramatically accelerating the learning curve for teams new to Industrial DataOps.
For an idea of what this looks like, watch Aron Semle’s recent highlight video where he demonstrates these MCP configuration tools.
Wrap Up
AI is transforming manufacturing, but transformation requires infrastructure. HighByte Intelligence Hub version 4.3 delivers that infrastructure, enabling enterprises to deploy highly available Industrial DataOps solutions, deepen integration with modern data platforms, and start innovating with industrial AI.
Beyond the features highlighted here, version 4.3 includes a new system dashboard for monitoring hub health, improvements to OPC UA path-based identifier mode, enhanced Oracle Database CDC capabilities, PostgreSQL performance optimizations, PI System query pagination, MQTT topic browsing, and much more.
AI is here to stay, and version 4.3 equips industrial enterprises to leverage it practically, without losing focus on the core capabilities that keep operations running.
To learn more, please check out these additional resources:
- Read the product announcement.
- Explore the release notes for details on all new features and fixes.
- Request a free trial or log in to your existing account to test and deploy the software in your unique environment.
