As AI reshapes productivity, data development is undergoing a quiet yet profound transformation — we are no longer satisfied with "AI writing code"; we expect it to truly understand the business, comprehend workflows, and autonomously execute tasks.
Today, we are excited to announce that DataWorks Data Agent is now generally available. As our newest agent built for the AI era, it deeply integrates the cognitive capabilities of foundational models with your enterprise data assets, supporting natural language interaction, autonomous task planning, and end-to-end closed-loop execution — propelling data development from a "human-driven system" to a new stage of "proactive system service."
DataWorks Data Agent is an AI agent we built for data development and governance scenarios. As a one-stop AI agent deeply integrated into the DataWorks platform, it breaks through the limitations of traditional Copilot "assistive suggestions" and officially upgrades to a digital workforce capable of independently planning and executing complex tasks.
The product comes with core capabilities including Agent, Code Programming Assistant, ChatBI, and Quick AI Actions, comprehensively covering the entire lifecycle of data integration, development, operations, governance, and analytics. Powered by advanced AI reasoning and natural language interaction, you can automate full-cycle tasks — from data integration and development to operations, quality governance, and data analysis — simply through conversation, delivering an efficient and trustworthy intelligent data development experience for your team.
In the coming release, DataWorks Data Agent will introduce custom Skills, enabling you to encapsulate proprietary data services and pipelines on demand. Our built-in Skills will expose standard calling interfaces for seamless integration with any third-party Agent. This will make professional data development capabilities truly "plug-and-play," achieving cross-platform flow and ecosystem-level collaboration.

DataWorks Data Agent -Chat Mode

DataWorks Data Agent-CLI Mode
DataWorks Data Agent spans the full data development lifecycle, making it possible to orchestrate the entire pipeline with a single prompt. Here's how it changes things across each stage:
Setting up data sync used to mean wrestling with data source connections, field mappings, partition strategies, and scheduling dependencies — all before you could move a single row. Now, you just tell the Agent something like "sync daily new MySQL data to Hologres," and it parses the intent, generates the configuration, and gets it running. The whole setup is generated by the agent, and you don't need to be a sync expert to get it right.
Building an ETL pipeline used to be a high-barrier task — you needed deep expertise in data modeling, SQL development, dependency management, testing, and deployment. It was slow, hard to standardize, and knowledge rarely carried over between projects.
With DataWorks Data Agent, that barrier is gone. Hand it a requirements document and say "build the ads layer for live-stream product transaction data." The Agent takes care of the heavy lifting — requirement analysis, node creation, code generation, dependency configuration, and testing — all the time-consuming development work is handled by AI. You just review and approve.
Note: Critical operations — such as publishing to production environments — still require manual operation. DataWorks Data Agent follows a human-in-the-loop model, ensuring you stay in control of every key decision.
What used to demand senior-level data engineering skills can now be done by anyone who can describe what they need.
Quality monitoring used to be a reactive, multi-step chore: find the table, inspect the fields, review the SQL, configure rules, test, save — and by the time you're done, days have passed. Now you just tell the Agent to "set up row-count quality rules for all tables starting with ods_." It analyzes field types, business semantics, and data importance on its own, then recommends and applies the right monitoring rules. Everything is fully auditable, and governance shifts from reactive cleanup to proactive control.
Finding the right table or tracking down who changed what used to mean flipping through lineage graphs, pinging colleagues, and digging through logs — easily 30 minutes or more per lookup. Now you just ask "which table has user shipping addresses?" or "who last modified this wide table?" and get results with full lineage and change history in seconds. It's like having a conversation with your data.
When a pipeline task fails, the old workflow meant manually pulling logs, tracing dependencies, and comparing trends — typically an hour or two before you even knew what went wrong. Now the Data Operations Agent does all that automatically: it gathers task logs, instance details, operation history, task code, and recent run status, then generates a diagnostic report with a recommended fix. You review and approve, and the Agent executes the repair. The entire process takes under five minutes.
Analytics used to follow a familiar bottleneck: business asks a question, an analyst interprets it, explores the data, writes queries, builds charts, and iterates — a cycle that typically takes one to three days. Now you just ask "how did sales compare YoY by region last month?" or "which category lost the most users?" and the Agent figures out the metric definitions, runs the query, and generates the chart — all in under a minute. It supports multi-turn exploration too, so you can drill down conversationally without writing a single line of code.
The most important topic for DataWorks Data Agent — security. Security and controllability are not add-on features; they are the first principle for Data Agent to be trusted by enterprises as a digital employee.
Two core security features:
Four security design principles:
In one sentence: Intelligence ≠ Uncontrolled. Getting security right is the foundation of DataWorks Data Agent's trust by enterprises.
To make it easy for both enterprise users and individual developers to get started, we are introducing a "Seat Fee + Pay-as-You-Go" hybrid pricing model, balancing experience and cost control.
| Tier | Target Users | Core Features | Bonus Token Quota (Limited-Time) | Price |
|---|---|---|---|---|
| Trial | Personal trial | Basic AI features | 500KTokens/month | Free |
| Personal | Independent developers | Basic AI features | 2M Tokens/month | $10/month |
| Team | Enterprise teams | Full features | 6M Tokens/seat/month | $20/seat/month |
💡 Your seat fee comes with far more value than it costs. The included Token quota significantly exceeds the price of the seat itself — giving you generous headroom to explore, build, and iterate before any additional charges kick in.
🔔 Limited-time bonus — subject to change. The Token quotas listed above are part of a promotional offer and may be adjusted at any time. Once your included quota is exhausted, billing automatically switches to pay-as-you-go. Upgrade your tier or add seats anytime for higher quotas.
👉 Activate DataWorks Data Agent Now: Activate now or view official documentation.
If you are existing users of dataworks,get started with DataWorks data agent here.
Hologres CLI & Skills: Agent-Ready Infrastructure for Smart Data Warehouse Ecosystem
17 posts | 0 followers
FollowAlibaba Cloud Big Data and AI - April 13, 2026
Alibaba Cloud Community - September 30, 2025
5927941263728530 - May 15, 2025
Alibaba Clouder - February 20, 2020
ApsaraDB - January 15, 2026
Alibaba Cloud Big Data and AI - January 21, 2026
17 posts | 0 followers
Follow
Big Data Consulting for Data Technology Solution
Alibaba Cloud provides big data consulting services to help enterprises leverage advanced data technology.
Learn More
Big Data Consulting Services for Retail Solution
Alibaba Cloud experts provide retailers with a lightweight and customized big data consulting service to help you assess your big data maturity and plan your big data journey.
Learn More
Realtime Compute for Apache Flink
Realtime Compute for Apache Flink offers a highly integrated platform for real-time data processing, which optimizes the computing of Apache Flink.
Learn More
Cloud Migration Solution
Secure and easy solutions for moving you workloads to the cloud
Learn MoreMore Posts by Alibaba Cloud Big Data and AI