Metad Analysis Cloud, an agile data analysis platform based on cloud computing, integrates multidimensional modeling, metric management, and BI presentation, aiming to provide users with an efficient and convenient data analysis experience. We are pleased to announce that the latest version 2.3 of Metad Analysis Cloud has been officially released. This update not only optimizes platform performance but also brings several exciting new features, particularly a comprehensive upgrade to Copilot Command and Roles.
Major Breakthroughs in Copilot
Copilot Command is an innovative way for users to execute AI functions. By entering commands along with corresponding prompts, the Copilot Agent will invoke large language models (LLMs) and run relevant functions to complete various data analysis tasks. The new version 2.3 of Metad Analysis Cloud has made significant improvements in this feature:
Introducing LangChain Agents Support
Metad Analysis Cloud is gradually transitioning the implementation of all Copilot Commands to the LangChain Agent method, making command execution more intelligent and flexible. The addition of LangChain Agents enhances data processing and analysis efficiency and accuracy.Enhanced Copilot Command Agents
The new version includes several enhancements to Copilot Command Agents:- Dimension Members Retriever: This feature dynamically retrieves and acquires key information of dimension members, helping users better understand data dimensions.
- Few Shot: By retrieving similar examples, it provides more precise templates, making data analysis more aligned with actual needs.
- Suggestion: Provides dynamic prompt completion functionality, helping users quickly generate high-quality analysis commands.
Introducing Business Roles
The new version also adds the Business Roles feature to the AI Copilot. Users can switch business roles during use, allowing for more precise business analysis in different scenarios. Whether it's market analysis, financial reporting, or operational management, different business roles will help users achieve personalized data analysis experiences.
Each business role has corresponding examples. When the Copilot executes an Agent, it selects the examples most relevant to the current task to send to the LLM, ensuring more accurate results. This feature enhances the specificity and practicality of the analysis, significantly improving user work efficiency and data insights.
Common Roles
Below is a list of common business roles categorized by different technical fields:
| Name | Title | Title (CN) | Responsibility Description |
|---|---|---|---|
| financial_analyst | Financial Analyst | 财务分析师 | Analyzing financial data, forecasting, budgeting, and financial reporting. |
| supply_chain_manager | Supply Chain Manager | 供应链经理 | Overseeing supply chain operations, logistics, inventory management, and supplier relationships. |
| marketing_specialist | Marketing Specialist | 营销专员 | Developing marketing strategies, executing campaigns, and analyzing market trends. |
| data_scientist | Data Scientist | 数据科学家 | Analyzing complex data sets, developing machine learning models, and deriving insights. |
| it_support_manager | IT Support Manager | IT 支持经理 | Managing IT support teams, resolving technical issues, and ensuring IT infrastructure stability. |
| hr_specialist | Human Resources Specialist | 人力资源专员 | Managing recruitment, employee relations, training programs, and HR policies. |
| operations_manager | Operations Manager | 运营经理 | Overseeing daily operations, process improvement, and operational efficiency. |
| product_manager | Product Manager | 产品经理 | Planning and overseeing product development, market research, and user experience. |
| customer_service_manager | Customer Service Manager | 客户服务经理 | Managing customer support teams, ensuring customer satisfaction, and handling complaints. |
| compliance_officer | Compliance Officer | 合规专员 | Ensuring compliance with laws and regulations, conducting audits, and implementing compliance programs. |
These roles cover common positions in various technical fields within enterprises, such as finance, supply chain, marketing, data science, information technology, human resources, operations, product management, customer service, and compliance. Each role's responsibility description helps clarify its specific role in business analysis and decision-making processes.
Other Updates
- GPT-4o supports the latest GPT-4 Omni model, providing faster and stronger natural language processing capabilities.
Architecture Planning
For the architecture planning of Copilot agents, we divide the agents into three levels: Command Agent, Business Roles, and Multi-Agent Collaboration.
- The Command Agent is responsible for executing user-input commands for single tasks.
- Business Roles provide examples and prompts based on users' business needs.
- Multi-Agent Collaboration coordinates the work between different Command Agents to achieve more intelligent and efficient data analysis.
We have completed the upgrade of some Command Agents and the functionality of Business Roles. The functionality of Multi-Agent Collaboration is currently under development, so stay tuned.
Conclusion
The release of Metad Analysis Cloud version 2.3 marks a new stage in data analysis. By introducing LangChain Agents and Business Roles, we aim to provide users with more intelligent, efficient, and personalized analysis tools. Whether you are a data analysis expert or a business decision-maker, Metad Analysis Cloud will be an indispensable assistant.
Upgrade to Metad Analysis Cloud version 2.3 now to experience a new way of data analysis and embark on a new chapter of data-driven decision-making.