Zoho introduces trusted business intelligence layer in Zoho Analytics

Zoho has announced a major evolution of Zoho Analytics, its business intelligence platform, introducing a robust intelligence layer for business insights and AI-driven action that extends the platform's established analytics and reporting capabilities. By seamlessly bridging data, business context, and action, the upgraded platform brings a high level of trusted intelligence to a business's apps and workflows.

"The business intelligence industry is undergoing a fundamental transition from traditional human-driven analysis to intelligent orchestration," said Clarence Rozario, Director of Product Management, Zoho BI and Analytics Platform. "The true promise of Agentic AI lies in executing complex workflows with absolute precision. However, to trust an AI agent with operational decisions, it must inherently understand an enterprise's unique metrics, hierarchies, boundaries, and business vocabulary. By engineering native semantic and ontology layers into Zoho Analytics alongside open standards, we are providing this critical structural context. We are transforming the platform into a centralised intelligence layer that empowers organisations to bridge the gap between visualising data and confidently automating business outcomes."

Multi-agent Layer Built on Business Context

The upgraded Zoho Analytics empowers businesses with a platform that analyses their data, interprets it, and enables AI agents to execute tailored workflows, all in one package. 

The latest version of Zoho Analytics introduces new ontology and semantic layers that help AI interpret data in the context of a business's operation. The ontology layer maps a company's organisational structure, while the semantic layer acts as a business glossary defining key metrics. Together, they ensure AI understands specific business contexts to deliver accurate, actionable results.

This deep understanding powers Ask Zia, Zoho's conversational BI super-agent. Built on a multi-agent architecture, Ask Zia orchestrates six specialised sub-agents (covering data analysis, modelling, visualisation, data preparation, coding, and SQL generation) to handle complex requests from a single prompt. Users can also build custom agents using this same knowledge base, which can connect internally or externally via Model Context Protocol (MCP). Ultimately, these agents go beyond analysing data to directly triggering automated workflows. Powerful workflows can now be designed through a workflow builder in Zoho Analytics. The upgraded platform gives businesses an all-in-one solution to interpret data and execute tasks autonomously.

Crucially, Zoho Analytics' native semantic and ontology layers eliminate the need for a separate cloud data warehouse. Users can build and edit multiple semantic models without having to write any code.

To enhance this intelligence layer further, 30 new data connectors are being added to Zoho Analytics in addition to the available 500-plus options. These new connectors expand Zoho integrations by adding support for Zoho ERP, Zoho Spend, Zoho FSM, Zoho Shifts and Zoho Voice and also expand interoperability with third-party services like SAP, PayPal, Adobe Commerce, Twilio, Notion, Intercom, Supabase, Amazon DocumentDB, and AlloyDB, bringing relevant data into Zoho Analytics for analysis and insights.

Furthermore, to prevent vendor lock-in, the platform supports Apache Ossie (an open-source standard that acts as a universal format for business logic) which allows organisations to define their metrics and dimensions once, and share them seamlessly across different BI tools and AI agents. This ensures that a company’s semantic definitions remain completely portable and never locked inside the Zoho ecosystem.

Portable Intelligence for Agents, Apps, and Workflows

For enterprise IT teams and developers, the intelligence is designed to be highly portable and extensible outside the Zoho ecosystem across external agents, workflows and applications.

The headless BI model allows developers and agents to build agentic solutions over the Zoho Analytics MCP server, available in remote and local options, with tools for data analysis, data preparation, and visualisation. A new command-line interface (CLI) is included, and users can work with Zoho Analytics from inside Claude through a Claude plug-in.

Through the embedded AI system, organisations can also embed agents in their own apps, portals, and products with white labelling. Permissions and governance remain intact wherever the agents run.

Built-In Governance and Oversight

As the tech ecosystem is beginning to accommodate AI regulations better, Zoho Analytics intuitively comes equipped with strict governance and clear oversight into each step of its process to help manage data access and cost, reduce any chance for errors, and maintain control over how information is used by agents.

Zoho Analytics gives administrators visibility into, and control over, how agents operate across the organisation. The Data Catalog brings together data, metrics, analytics assets, and models, and offers lineage, impact analysis, broad search options, and management of data and access policies. Through AI observability, administrators can monitor which agents ran, who used them, what they accessed, and how they answered queries. AI usage and cost tracking across teams, along with administrative controls for Ask Zia, helps organisations monitor and manage AI adoption. Ask Zia also provides organisations the choice between Zia’s own large language model or their own frontier model.

Zoho Analytics also provides sandbox and deployment management capabilities through which teams can develop and test analytics artifacts and agents in a sandbox, publish to production when ready, and roll back to previous versions if necessary.

Additional Capabilities

Zoho Analytics offers more than 50 new capabilities beyond its agentic features, including more than 10 new visualisations, custom visuals, and pixel-perfect reports; predictive tools such as scenario analysis, a correlation matrix, and a VARX forecasting model; a data science studio with Jupyter notebooks, PMML import, and expanded analysis types; and Ask Zia on mobile with an upgraded interface.