Ask Zia for Data Analysts and Business Users
Data analysts and business users don't have to sift through many reports to understand the key insights. Ask Zia provides contextual insights that not only explain what is happening in your data but also highlight the underlying reasons behind those trends and anomalies. Ask Zia, the super agent simplifies the process of creating reports and dashboards.
- Visualization Agent: Understands analytical requirements expressed in natural language and generates relevant charts, reports, dashboards, and summary widgets using the available data. It helps users explore business metrics, compare performance across dimensions, and visualize trends without manually configuring each report.
- Analysis Agent: Examines data to identify trends, patterns, correlations, and anomalies. It helps users investigate changes in business performance, understand the factors contributing to those changes, and derive contextual insights to support informed decision-making.
Supported Plans
The Analysis Agent and visualization agent capabilities are available across all Zoho Analytics plans, including the Free, Basic, Standard, Premium, and Enterprise plans.
Invoking Ask Zia
Access Ask Zia, the super agent, from either the left navigation panel or the workspace navigation bar. Selecting Ask Zia opens the conversational interface, where you can enter prompts, configure the relevant context, and interact with the super agent to perform supported tasks.
Add Context for Ask Zia
An AI agent is only as effective as the context available to it. Ask Zia allows you to configure the relevant data context to help the agent better understand your queries and generate more accurate, relevant responses.
- To configure the context, click the Add context drop-down menu on the left side of the prompt bar.
- The Context field allows you to choose which one should be used,
- Table
- Semantic Model
Tables
The Tables option allows you to select specific tables and columns that Ask Zia can use when processing your queries.
- Select Tables under the Context section.
- Use the search bar to find specific tables or columns. You can also use the filter option to narrow down your selection.
- Expand a data source or table to view its available columns.
- Select the checkboxes next to the tables or columns you want to include in the context. Use Select All Tables to select or deselect all available tables.
- Click Apply to save your selection and apply the context to Ask Zia.
Only the selected data is included in the configured context, helping Ask Zia focus on the relevant information when responding to your queries.
Semantic Model (Preview)
The Semantic Model option allows you to provide context through a business-oriented representation of your data. It helps Ask Zia interpret queries using defined business concepts, metrics, and relationships rather than relying solely on raw table and column names.
Select Semantic Model under the Context section to configure the semantic context available for your queries. The available configuration options depend on your semantic model setup.
Click Apply to confirm your selected semantic model and apply it as context for Ask Zia.
Vizualization Agent
The Visualization Agent supports report and dashboard creation in Zoho Analytics, enabling users to explore data and present business information visually. This section covers the different ways to create reports and dashboards using Ask Zia.
Creating Reports
Create insightful reports by describing your requirements. Ask Zia understands the analytical intent, generates suitable reports, applies relevant filters or groupings, and recommends follow-up views or questions, thereby making report building faster, smarter, and accessible to all users.
- Access the Ask Zia Portal from the side navigation panel. Ensure to enable the conversation mode.
- In the prompt section, describe what report you want to create. Examples: Show me the region sales trend report. Refer to the Constructing questionsarticle to learn how to frame effective prompts for generating reports and visualizations.
- To make the conversation more easier, Zia auto suggests a few questions based on your data. Click the more icon and then select the questions you want to ask. Zia will get you the answers.
- Click the Save icon to add the reports to the workspace.
- Specify the View Name and choose the Folder Name where it should be saved.
- Click Save.
- The Change chart type option allows you to change and choose the suitable chart type.
- Zia insights provides a detailed summary of key insights in the report.
- The Edit option allows you to modify the report settings such as axes, legends, and more.
- The Add to Dashboard option allows you to include the reports in an existing dashboard in the workspace.
- The Export icon allows you to download the generated report.
Agent Thinking Process
The Agent Thought Process option allows you to view the reasoning steps Zia follows while generating a response. It provides transparency into how the AI interprets your question, understands the data model, and arrives at the final output.
Zia displays the intermediate steps it performs during the analysis. This can include identifying relevant tables and columns, interpreting the metrics and dimensions mentioned in the query, determining the appropriate aggregation, and constructing the report or analysis needed to answer the question.
For example, when a user asks for month-wise sales, Zia first examines the data model to locate the tables that contain sales and date information. It then determines how the sales metric should be aggregated and how the date column can be grouped by month. Based on this understanding, Zia generates the report showing sales by month.
This can be useful for validating the logic behind the response and gaining deeper insight into the analysis performed.
Key Highlights
Key highlights summarizes the most important insights from a report or chart, helping users quickly understand the key patterns in the data.
The insights are generated using the chart metadata, including the metric, dimensions, and aggregated values already present in the visualization. No additional columns or external data are introduced during this process, ensuring that the insights strictly reflect what is shown in the chart.
Based on this metadata, the system identifies relevant analytical observations such as contribution, comparisons between entities, trends over time, and forecast-related observations where applicable. These analytical findings are then sent to a Large Language Model (LLM) to generate a clear and concise natural language summary.
This enables users to instantly grasp the key takeaways from the chart without having to analyze the visualization manually.
Creating Dashboards
Whether you’re building dashboards from scratch or enhancing existing dashboards, Ask Zia understands your layout preferences, content needs, and visual goals and delivers a personalized experience every step of the way.
From the Ask Zia Portal
- Access the Ask Zia Portal from the side navigation panel.
- In the prompt section, provide the details about the metrics to be used for the dashboard creation. Example: Create a dashboard using sales metrics.
- Click Save to add the reports to the workspace.
- Click Open in a new tab to modify the dashboard settings.
- Specify the details on customizing the dashboards in the prompt section.
- Layout: Specify how the charts and widgets are to be arranged on the dashboard, whether single, double, or custom column layout.
- Switch Themes: Customize the dashboard appearance by specifying the preferred theme. Example: Switch to Dark Theme.
- Add or Remove Views: Include or exclude specific charts, KPIs, or visualizations as needed to refine your dashboard. Example: Add sales by region report.
From the Dashboard Designer
- Click Create > Dashboard on the side navigation panel.
- In the prompt section, provide the details about the metrics to be used for the dashboard creation. Example: Create a dashboard using sales metrics.
- Prompts executed here are reflected in real time on the dashboard canvas.
- Customize the dashboard design as per your preferences. Refer to the above section to learn more.
Explore Data Insights with Ask Zia
The conversational interface in Ask Zia allows users to explore data using simple language. It offers a closer lens into what has happened, why it happened, and what’s likely to happen along with recommendations to guide your next steps.
Analysis Agent
Ask Zia enables users to go beyond report generation by asking follow-up questions and exploring their data from different analytical perspectives. Users can perform descriptive, diagnostic, and prescriptive analysis to understand business performance, investigate trends, and identify potential actions.
Descriptive Analysis
Descriptive analysis helps users understand what happened by summarising historical and current data. Ask Zia can answer questions about business performance, identify trends, and highlight patterns in the data through reports and visualisations.
Users can ask descriptive questions to obtain an overview of key metrics and understand how performance changes across time periods, categories, or business segments.
Example questions:
- What was the total sales revenue this quarter?
- Which products generated the highest revenue?
- How did monthly sales perform over the past year?
- Which region recorded the highest number of sales?
Diagnostic Analysis
Diagnostic analysis helps users understand why something happened by examining data in greater detail. Users can ask follow-up questions to investigate changes in performance, compare business metrics, and identify factors that may have contributed to a particular outcome.
Ask Zia helps users explore relationships and patterns in the available data to investigate potential reasons behind business results.
Example questions:
- Why did sales decline this month compared with the previous month?
- What factors contributed to the decrease in revenue?
- Why did one region perform better than another?
- Which product categories contributed most to the decline in sales?
Recommendations
Prescriptive analysis helps users explore what actions they could take next based on the available data and identified trends. Users can ask questions about potential strategies, improvements, and opportunities to support business decision-making.
Based on the available data and analytical capabilities, Ask Zia can provide recommendations or suggest areas that may require further investigation. These suggestions should be evaluated against business objectives and constraints before implementation.
Example questions:
- Which product categories should we focus on to improve revenue?
- What actions could help improve sales in underperforming regions?
- Which customer segments should we prioritise for a marketing campaign?
- How could we improve campaign conversion rates?
Creating Slideshows
Convert reports and dashboards into compelling, narrative-driven slideshows with Ask Zia.
- Access the Ask Zia Portal from the side navigation panel. Ensure to enable the conversation mode.
- In the prompt section, describe the slideshow details you want to create. Examples: Construct a slideshow using sales metrics.
- Click the drop - down icon to view the reports used in the slideshow. you can easily add or remove views from the slideshow.
- Click Open in new tab to view the slideshow.
- Rename the slideshow by specifying the preferred slideshow name in the prompt.
Provide Feedback on Zia’s Responses
Zia Insights allows you to provide feedback on the responses it generates. Your feedback helps improve the accuracy and relevance of future insights.
Each time Zia generates a response, you can rate its quality using the thumbs up (👍) or thumbs down (👎) icons.
- Thumbs Up – Select this if the insight or explanation was helpful and accurately addressed your query.
Thumbs Down – Select this if you’re not satisfied. This will open a follow-up box asking for the reason, which helps improve Zia’s future performance.
Providing feedback helps improve Zia’s ability to generate more accurate, relevant, and useful insights over time.
