ADBAgent Getting Started Guide
ADBAgent is the built-in AI data analysis assistant in AnalyticDB for MySQL. You can perform professional data analysis using natural language instead of writing SQL, and save your analysis workflows as reusable Skills and business scenarios. This topic describes the features of ADBAgent and walks you through the entire process from environment setup to data analysis.
What is ADBAgent
ADBAgent is powered by the AnalyticDB for MySQL proprietary AI data analysis engine, which deeply understands database schemas and business semantics. It delivers an end-to-end workflow from natural language interaction to intelligent analysis, Skill creation, and scenario reuse.
Use cases
Business operations analysis: Data analysts configure analysis scenarios in ADBAgent and bind core tables and Skills. Business users can then self-serve routine data queries in natural language. Results can be scheduled and delivered to DingTalk or Feishu.
Cross-department data collaboration: Centralized data scenarios and Skills ensure consistent metrics across teams. Shared scenarios prevent redundant work, and full traceability simplifies metric discrepancy troubleshooting.
Automated business decision-making: Integrate ADBAgent with OpenClaw through a Skill to trigger recurring analysis tasks and execute optimization strategies automatically, closing the loop from data insight to action.
Features
Feature | Description |
Natural language interaction | Supports multi-turn conversational data analysis. ADBAgent automatically interprets user intent, generates SQL, and presents results. |
Three-layer data exploration | Deeply mines database schemas and business semantics to provide accurate data context for natural language queries. For more information, see Three-layer data exploration. |
Industry-specific analysis Skills | Includes built-in business metric analysis Skills for verticals such as finance and advertising. You can also generate custom Skills using natural language or download open-source Skills from the marketplace. For more information, see Industry-specific analysis Skills. |
Scenario orchestration | Allows you to orchestrate analysis Skills into reusable business scenarios for automated, recurring execution of high-frequency analysis tasks. |
Channels integration | Supports integration with third-party messaging tools such as DingTalk, enabling interaction with ADBAgent from anywhere. |
OpenClaw integration | Allows you to integrate ADBAgent into OpenClaw through a Skill. For more information, see OpenClaw integration. |
Core Skills
Three-layer data exploration
ADBAgent uses an on-demand, progressive strategy that standardizes data exploration into three layers, delivering full-pipeline data insights at minimal cost.
Layer | Purpose | Core actions | Problem solved |
Shallow: Schema overview | Builds a data overview map in seconds | Automatically parses column types, table sizes, primary key constraints, and physical storage strategies (partitioning, indexes). | Quickly answers "what data exists" and "how is it structured", establishing the physical foundation for subsequent analysis. |
Medium: Quality profiling | Deep-scans data distributions and potential risks | Performs statistical distribution analysis on numeric, string, and datetime columns. Identifies missing values, duplicates, outliers, and format inconsistencies. | Surfaces data quality issues early, quantifies data health, and prevents misleading conclusions drawn from dirty data. |
Deep: Semantic insight | Connects business meaning to execution performance | Infers business semantics for columns, discovers multi-table join logic, and evaluates query performance bottlenecks, access permissions, and sampling strategies. | Translates data characteristics into actionable business recommendations to guide report generation, model training, and SQL optimization. |
Core value: Simple queries get instant responses, while complex questions receive deep insights — making data analysis more accurate and efficient.
Industry-specific analysis Skills
In the advertising and attribution analytics domain, ADBAgent includes built-in industry-level Skills such as conversion rate analysis and campaign diagnostics. These Skills form a comprehensive diagnostic framework from macro trends to micro root causes to strategy recommendations.
Capability | Description |
Full-funnel multi-dimensional drill-down | Combined with data exploration, drills into key dimensions such as traffic channels, media types, operating entities, and campaigns, pinpointing issues at the right level of granularity. |
Conversion efficiency ratio model | Uses a proprietary "conversion share / click share" efficiency metric that eliminates volume bias. Objectively quantifies relative traffic quality to quickly identify high-potential sources and low-efficiency drains. |
Risk-tiered alerting | Automatically segments sources into "high-risk drain" and "high-efficiency scale-up" tiers based on efficiency thresholds, providing clear red/green decision signals for budget allocation. |
Cross-dimensional root cause analysis | Performs multi-dimensional cross-analysis to pinpoint hidden "high-risk combinations", distinguishing between systemic failures and localized structural anomalies. |
Closed-loop strategy output | Follows a "diagnose → attribute → recommend" logic to produce actionable optimization steps, completing the full loop from data insight to ROI improvement. |