Zoho Integrates Native Semantic and Ontology Layers in Latest Analytics Platform Update

Zoho has launched a major restructuring of its analytics platform, introducing native semantic and ontology layers designed to bridge raw data sources and AI-driven applications. According to company executives, the release positions the platform as a trusted intelligence layer capable of serving both human operators and autonomous digital agents without requiring a separate third-party data warehouse.

Architectural Pillars and Context Layers

The updated software is organized around four core pillars: trusted data and business context, agentic analytics and actions, intelligence everywhere, and governed AI. To establish robust business context, the platform utilizes two distinct layers positioned above raw data integration points. An ontology layer defines canonical business entities and their interrelationships, while a semantic layer translates those relationships into business-friendly metrics, descriptions, sample questions, and corresponding SQL queries.

These structures can be generated manually or through integrated AI assistance. Furthermore, the architecture supports compatibility with the Open Semantic Interchange specification, a vendor-neutral standard designed to facilitate seamless metadata exchange across different data infrastructures.

Expanded Agentic Capabilities and Deployment Flexibility

The natural-language interface, AskZia, has evolved into a multi-agent orchestration system. It incorporates pre-built sub-agents dedicated to specific tasks such as data analysis, visualization, preparation, modeling, coding, and workflow execution. Additionally, a new custom agent builder allows organizations to construct specialized agents grounded on specific datasets or semantic models.

To support flexible deployment, the platform emphasizes composability. Enterprises operating on external data platforms can utilize zero-copy connectors to pull metadata while leaving core data in place. Organizations can also adopt individual layers independently, integrating them with existing enterprise architectures.

• Native semantic and ontology layers built directly into the platform
• Multi-agent orchestration supporting custom agent creation and workflow execution
• Composable architecture compatible with external data storage and open standards