“Business Intelligence” (BI) is BIG stuff. It means collecting, analyzing, and presenting data to help organizations make informed decisions, improve performance, and gain strategic insights through reporting and analytics. It is easy to define, but hard to deliver because, for most businesses, data are scattered across applications, systems, departments, private clouds, and public clouds. Meanwhile, executives and data analysts need tools that aggregate and analyze data in “near real time” and present it in formats that fit their immediate needs, whether that’s a set of charts, a table or a short response to a natural-language query.

In a series of posts this year, my colleague Amy Stapleton recognized Zoho’s ongoing effort to “democratize BI.” In January she took note of Zoho’s AI Blueprint, and how it accommodates efforts to match model size and purpose to each company’s budget. In June she provided specific examples of instances where Zoho has integrated fit-to-purpose AI models into use cases that enhance productivity and provide insights. The latest version of Zoho Analytics, announced today (September 12), is the culmination of ongoing efforts to give managers, departmental execs, and analysts the tools they need to get practical questions answered or actions taken in response to natural language input.

“The Analytics Platform for All”

Zoho Analytics overcomes the challenges of democratizing BI by taking an end-to-end approach with four key focus areas:

1. Data Management: enabling the “data experts” among a company’s staff to create and manage complex data pipelines.
2. Generative AI-infused analytics: providing both analysts and data users to use their own words to get answers to questions regarding key business drivers.
3. Data Science and Machine Learning (DSML): acting as a studio for data scientists and “citizen data scientists” to create relevant models automatically (AutoML) or from scratch.
4. Platform Extensibility: simplifying the process of building connectors to popular CRM, ecommerce, video and data management systems.

The Data Management Hub is the foundation of the entire suite. It is the result of “deep integrations” of data from a multiplicity of applications and tools. The word “deep” is used advisedly. It takes advantage of “stream analytics” tools (Cloud Pub/Sub, Kafka, and PubNub), plus 25 new data connectors to extract data from popular business apps (including Oracle Netsuite, Microsoft Dynamics 365, and many others), and many more.

Zoho has built a common data model for the ingested data and can, then automatically generate predefined reports. There are also tools for IT-managers to define and manage complex data pipelines using existing tools and features. They can create end-to-end pipelines using a visual builder; custom “transforms” and ML models with Python Code Studio or by using Ask Zia as a natural language user interface; and orchestrate data pipelines with Zoho Flow. As a final step, the Data Management Hub includes a “Metrics Layer” with resources for data modeling, a metrics store, access control, and “headless BI” necessary for trustworthy quality management and administration.

GenAI Supports Self-Service Report Generation and Trigger Actions

Business analysts, data analysts and departmental executives share a need for near-real time business reports that get to “Why?” they see the results being displayed. Zoho Analytics delivers with GenAI-infused Smart Key Driver analysis. For example, while displaying a timeline of sales and profitability, it provides additional context, such as impact of changes in ad spending on a particular channel. In many instances, the query-and-response take the form of a conversation with Ask Zia, Zoho’s GenAI-infused co-pilot.

Speaking of Ask Zia, Zoho Analytics brings it some major enhancements. For one thing, it supports more languages (and will continue to add more). More importantly, its insights can trigger specific actions and even generate new use-case specific, custom models. What’s more, these insights can be shared with other analysts and executives across instant messaging channels.

GenAI plays an important role in popularizing BI and its results by automatically evaluating existing reports, and suggesting or generating new metrics and reports when it detects a need. These fuctions are customizable to suit the needs of specific executives or analysts. For those wondering, Zoho Analytics also supports refinement of responses for accuracy by employing Retrieval Augmented Generation (RAG). It is powered by a “seamless” integration with APIs from OpenAI, and also supports a “Bring-Your-Own-Key” (BYOK) approach to your LLM of choice.

Tools for Generating, Managing and Publishing new Models

The final two focus areas are DSML Studio and “Platform Extensibility.” This is where Zoho’s long-standing development of AutoML comes into play. Data scientists and “citizen data scientists” can create new machine learning models automatically, without coding. They can then train, test, compare, and manage models as needed to make sure they are useful and effective. Then, BI developers and data engineers will find that Zoho’s composable approach makes it easy to build connectors or publish their BI resources to marketplaces like HubSpot, Zendesk, Azure, Shopify, LinkedIn, YouTube, MongoDB, and others.

The latest version of Zoho Analytics marks the realization of long-standing efforts to democratize BI. Zoho calls it “end-to-end,” but it is also a “bottoms up”, full-stack approach. Great attention is paid to the details of bulletproofing the lower layer plumbing and connectors as well as data layers that are trustworthy and “open” to work well with easy-to-use report generators and natural language queries. Kudos to Zoho.

Categories: Intelligent Assistants, Articles