MaxCompute - Agentic ecosystem toolkit launched
Jul 13 2026
MaxComputeContent
Target customers: 1. Data analysts and data scientists: individuals and teams who want to quickly perform data exploration, data import and export, visualization, and predictive analysis using natural language, SQL, or AI agents. 2. Big data development and O&M teams: scenarios such as building enterprise-grade lakehouses, hybrid SQL/Python development, dbt integration, and table and snapshot backup management. 3. AI application and Agent developers: enterprises and developers who want to seamlessly connect enterprise-grade big data capabilities to large language models to enable automated data analysis and intelligent O&M. . New Features and Specifications: The MaxCompute Agentic ecosystem toolkit covers the full spectrum from desktop interactive analytics to AI Agent automated integration. Key features: I. AI Data Exploration Client-Integrates Data Agent, SQL IDE and Multimodal Data Management AI Query: Business users can perform self-service analysis without SQL expertise. Ask a business question to get data visualizations, insights, and exportable reports. Conversational data analysis provides full visibility into query and data processing stages with a complete evidence chain. Expert analysis modes, such as time series forecasting, anomaly detection, and multi-window trend analysis, can be invoked on demand. Semantic packages and a SQL knowledge base further ensure accurate alignment with business definitions. AI Coding: generates SQL and Python code that can be used for production based on user requirements and data schema information. SQL queries come with built-in analysis for data visualization, job performance, and result quality. Python UDF development covers the full workflow from building and debugging to release. Multimodal data management: designed for storing, processing, querying, and using enterprise multimodal datasets. Multimodal data stored in formats such as Blob gains both the clarity of table-based management and the ease of file-system-style operations. Drag-and-drop batch file import, and online preview and download of audio, video, images, and text are also supported. Desktop application: supports macOS 12.3+ and Windows 10+. Launch with one click and instantly connect to massive data in the cloud. II. MCP Server (MCMCP): Securely connect your Agent to MaxCompute Built on the MCP protocol, it encapsulates metadata, computing, and table management as Agent-callable tools, with support for natural language search, cost estimation, asynchronous execution, and read-only protection. It provides multiple authentication methods, including OAuth and AK/SK, and works with clients such as OpenClaw, Qwen Code, and Qoder. III. MaxCompute CLI: A command-line tool for scripts and Agents Invoked as aliyun maxc through Alibaba Cloud CLI, it supports running queries, managing jobs, browsing metadata, and uploading or downloading data with a single command. It can be easily embedded into enterprise automated data pipelines and intelligent scheduling workflows.
Help Document
https://www.alibabacloud.com/help/en/maxcompute/getting-started/maxcompute-client-maxcompute-query-studio-user-guide