💡 What Is Agentic BI?
Agentic BI (Agent-Driven Business Intelligence) is a new BI paradigm for the AI era. It embeds AI agents into the full data analysis lifecycle, enabling the system to understand, plan, and act—achieving true automation and intelligent decision support. In traditional BI, users must manually model, configure reports, write formulas, and interpret charts. In Agentic BI:- 🧠 AI agents become the core analysts: Just describe your business problem in natural language, and the agent understands your intent, constructs the analysis path, and executes it.
- 🔧 Agents can call toolsets: such as semantic modeling tools, indicator management tools, etc., to automate querying, calculation, interpretation, and recommendations.
- 🔁 Analysis is no longer a one-time result: it becomes a continuous dialogue, supporting follow-ups, refinement, and iterative thinking—realizing human-machine collaborative analysis.
🔧 Overview of Currently Available Features
🚀 Quick Start: 3 Steps to Activate Agentic BI
Step 1: Create a Digital Expert Workspace
Go to any Digital Expert interface and click “Create Workspace.” This space will host your semantic models, indicator system, business data, and agent conversation context—it’s the core container of Agentic BI.Step 2: Instantiate Toolsets
Within the workspace, click “Add Toolset” and choose from the available ones:- Semantic Model Toolset: Define business entities, dimensions, measures, etc.
- Indicator Management Toolset: Build and maintain key business indicators, define definitions, statistical rules, and business domains.
Step 3: Create an Agent and Bind Toolsets
Click “Add Agent,” choose a Digital Expert template (like “Financial Analyst” or “Market Insight Assistant”), and bind it to your configured toolsets. Once published, your Agentic BI agent is ready to go and can:- 📊 Create semantic models via conversation
- 📌 Define and refine indicators via natural language
- 📈 Analyze business using models and indicators
- 🧠 Perform multi-turn reasoning and generate suggestions
🔍 Live Demo: What Can Agentic BI Do?
Here are some example natural language interactions: Modeling:“Create a semantic model for sales orders with fields: customer, product, sale date, and sale amount.”Indicators:
“Add a indicator called ‘Monthly YoY Growth Rate’ to calculate the sales amount year-over-year percentage change.”Analysis:
“Show me the sales trend for East China by product line over the past 3 months, displayed monthly.”Planning:
“We’re preparing a gross margin analysis—please draft a indicator evaluation and analysis path.”
🤝 Multi-Agent Collaboration: Role-Based Analysis
Add multiple digital experts (agents) to a project for role-based collaborative analysis:- Market Strategy Assistant → User data analysis and segmentation
- Sales Forecast Expert → Trend prediction and planning recommendations
- Financial Analyst → Cost structure optimization and P&L analysis
🧩 Use Case Examples
✅ Why Choose Agentic BI?
🔍 Agentic BI vs ChatBI (Text2SQL)
✅ Example
ChatBI Question:“Query monthly sales for 2024.” 👉 Executes SQL and shows table/chart. No background understanding or indicator explanation.Agentic BI Question:
“How is our sales trend this year? Which product line contributed most?” 👉 Agent uses the semantic model to identify “sales,” applies indicator definitions, understands “trend” as time-series analysis, and supports follow-up questions like “which channel had more impact.” Fully contextual and explainable.
🧠 One-Sentence Summary of the Core Difference:
- ChatBI is an enhanced query tool—translating natural language into SQL.
- Agentic BI is an intelligent analysis assistant—it understands your question, builds analysis paths, invokes tools, explains results, offers suggestions, and even drives the next step.
🔜 What’s Coming Next?
We’re developing the following features—stay tuned:- Dashboard Toolset: For drag-and-drop chart configuration and result presentation
- Data Governance Toolset: For consistency checks, field quality tracking, and permission control