This article introduces the Quick BI V6.2 upgrade that deeply integrates AI into enterprise data decision-making workflows.
Over the past two years, AI has rapidly integrated into BI scenarios; enterprises now expect more from BI than just "viewing data"—they require systems that can understand business operations, identify issues, and support decision-making. However, in practical implementation, the ability to comprehend specific business metric definitions, pinpoint the root causes of anomalies, and reliably generate reports that integrate into collaborative workflows remains critical to whether these AI solutions earn trust and see sustained use.
This upgrade to the "Smart Q" capabilities in Quick BI V6.2 centers on the goal of making AI truly deeply integrated into enterprise operations—ensuring it is trustworthy, usable, and manageable. It encompasses nine capability updates across six key areas: Smart Q Interpretation, Smart Q Reports, Smart Q Insights, Knowledge Base, Spreadsheet AI Functions, and Data Analysis Skills. The upgrade focuses on five strategic pillars—intelligent root-cause insights, precise business understanding, flexible collaborative workflows, intelligent text processing, and open ecosystem expansion—to better align the product with actual business scenarios and more effectively meet the needs of real-world enterprise operations.
Quick BI V6.2: Overview of New Capabilities
1. Intelligent Root Cause Insights: From "Presenting Data" to "Analyzing Causes"
Feature 1: Smart Q– Added "Smart Q Insights"
Application Scenarios
When business teams encounter metric anomalies—such as a sudden drop in sales—traditional methods struggle to rapidly pinpoint root causes across multiple dimensions and steps. This difficulty stems from a lack of automated tools capable of performing multi-step attribution and cross-period comparisons.
Feature Overview
-
Metric Insights: Conducting multi-dimensional, multi-step root cause analysis on changes in specific metrics.
-
Operational Insights: Analyzing insights across multiple metrics to construct a unified, holistic business perspective that comprehensively and accurately reflects the enterprise's operational status.
"Smart Q Insights" marks the evolution of Quick BI’s AI capabilities from "passive response" to "proactive insight"—the AI no longer merely answers user-posed questions but can actively initiate a comprehensive attribution analysis, providing more timely support for business decision-making.
2. Semantic Alignment: Enabling AI to Accurately Understand Enterprise and Individual Business Language
Over time, various enterprises and individuals have developed their own unique business logic and analytical methods. Simply integrating general-purpose large AI models into BI systems for inference—without aligning with the specific business context and analytical habits of the organization—often yields results that are inconsistent and prone to significant error. This upgrade addresses this issue from three perspectives.
Feature 1: Smart Q interprets the integration with the corporate knowledge base and analysis logic to generate data insights tailored to the enterprise.
Feature Overview:
Integrate with your corporate knowledge base—enabling AI to "understand" your business language.
-
One-click integration with the corporate knowledge base: A "Corporate Knowledge Base Search" feature has been added to the "Smart Q Interpretation" interface. Once enabled, the system automatically retrieves relevant business definitions and knowledge snippets based on the user's query, providing them as context for the AI's analysis to enhance the accuracy and business relevance of the interpretation results.
Define the analytical approach—have AI interpret data "your way."
-
Analysis Approach: Organization administrators can centrally configure standardized analysis approaches. Authorized members can select these directly when using "Xiao Q" for interpretation, and the AI will interpret the data step-by-step according to the preset analysis framework.
-
Content Templates: Supports the configuration of standardized report output structures—specifying required sections, key analytical focuses for each chapter, and output formats—to ensure that AI-generated interpretation reports are ready for immediate use in presentations.
Feature 2: Smart Q supports personal knowledge bases, enabling the reuse of personal analysis experience.
Use Cases
While corporate knowledge bases capture organization-wide analytical approaches and content templates, individual analysts often possess their own unique analytical habits, preferred lines of reasoning, and personal templates in actual practice.This type of knowledge, rooted in individual perspectives, is difficult to integrate into a unified corporate knowledge system and is unsuitable for sharing with other users; consequently, relying on personal experience necessitates repeated manual execution based on memory, preventing it from being learned or reused by AI.
Feature Overview
-
Personal Knowledge Management: After enabling the "Designate Analysis Approach" feature via entry points such as "Smart Q Report" or "Smart Q Interpretation," users can switch to their personal knowledge base to upload files and create custom analysis approaches and content templates; this personal knowledge is visible and accessible only to the user.
-
AI-Powered Learning and Synthesis: Upon uploading local files, the system automatically learns and distills your unique analytical approach and content templates.
-
Multi-scenario knowledge retrieval: Users can access "Smart Q" capabilities (such as Smart Q Insights and Smart Q Reports) at any time via the front-end interface to generate personalized analysis results.
-
Secure permission isolation: Personal knowledge bases are strictly isolated based on user levels, and permission verification mechanisms are implemented for scenarios such as collaborative editing to ensure that personal knowledge is not accessed or used without authorization.
Feature 3: Enhanced accuracy of Smart Q's interpretations
Application Scenario:
When business personnel make data-driven decisions and business judgments based on AI-generated interpretations, they often encounter issues such as data extraction biases and errors in calculating key metrics, which increase decision-making risks. Furthermore, the "black-box" nature of the analysis process makes it difficult to verify the credibility of the interpretations, severely limiting the application of AI-generated insights in business operations.
Feature Overview:
More precise data extraction & fully transparent execution process
-
Data Structure Governance: Dashboard data is transformed into standardized, structured data, ensuring that the data fed to the model for interpretation is clearer and more consistent, thereby eliminating data extraction bias at the source.
-
Intelligent Step-by-Step Processing Mechanism: Employs a multi-step strategy for invoking large models to effectively handle data volume limits, ensuring comprehensive interpretation even for complex dashboards.
- Added the "Smart Q Execution Process" information panel, clearly displaying the six key stages and the full interpretation workflow.
3. Flexible Collaboration: Integrating AI into Enterprise Workflows and Toolchains
The value of data insights is ultimately realized through their ability to integrate into an enterprise's collaborative workflows and existing ecosystem of tools.As AI analytics moves into the production phase, enterprises are demanding more in terms of proactive delivery, consistent results, and collaborative decision-making. This upgrade introduces three key capabilities that transform AI output from "passive triggering" to "flexible collaboration."
Feature 1: Smart Q Reports support pre-generated tasks.
Application Scenarios:
For report subscriptions targeting managers, data reports need to be pushed to leaders at a fixed time (e.g., 8:30 AM daily). The process may encounter the following problems:
- Before pushing a report, ensure the report content is accurate and permissions are correctly matched.
- Reports cannot be scheduled to refresh data on a timer; manual updates are required before each push, which is tedious and prone to errors.
- Reports with data row permissions are forced to refresh in real-time on every access, resulting in long loading times, unstable content, and severely impacting the reading experience.
- When multiple people subscribe to the same report, each triggers its own calculations, leading to inconsistent data and wasted computing resources.
Feature Overview:
A new "Pre-Generation Task" has been added to the report editing mode, which automatically updates report data on a scheduled basis.
- Scheduled Generation Configuration: Supports daily, weekly, and monthly scheduled triggering of pre-generated tasks, automating manual refresh processes.
- Data Permission Policy: Supports "Task Owner" (generating shared data based on editor permissions) and "Designated User" (generating independent data for each designated user, up to 30 people).
Pre-generated task management center
-
Task Management Entry: A new "Little Q Report Pre-generation Task" module has been added to the subscription management section, visible to users with report editing permissions.
-
Task Operations: Supports manual execution, pausing scheduling, resuming scheduling, transferring responsibility, viewing logs, and deleting tasks.
Feature 2: Smart Q Interpretation supports data interpretation subscriptions and pushes results to external channels such as DingTalk, Lark, WeChat Work, and email.
Application Scenarios:
Different business roles within an enterprise often have different needs for data interpretation, and subscription tasks need to simultaneously meet the needs of the following roles:
- Business personnel monitor core operational metrics (GMV, order volume, user activity) daily and require the latest business updates and data.
- The operations team regularly monitors the effectiveness of marketing campaigns (major promotion reviews, advertising ROI) and requires consistent, periodic data interpretation reports.
- Managers periodically report on team KPI achievement and require automated data insight delivery to reduce manual processing costs.
- Analysts have long-term, consistent analytical needs for specific dashboards and require proactive data interpretation rather than passive viewing.
Feature Overview:
The subscription feature delivers data analysis content to users on a scheduled basis.
- Supports periodic scheduled tasks: daily, weekly, monthly, and every N hours.
- Supports multi-channel push notifications: in-app messages, email, DingTalk, Lark, and WeChat Work.
- Supports flexible configuration: customize interpretation approach, interpretation scope, interpretation model, and enterprise knowledge base search.
Subscription support for personal and enterprise-level task management
-
Personal Task Management: Access "Understanding Subscription Task Management" via the subscription icon in the upper right corner. This allows you to start/stop, edit, and delete tasks.
-
Enterprise Task Management: Supports immediate sending, pausing scheduling, resuming scheduling, viewing logs, and batch operations.
Feature 3: Smart Q Reports support comments, and comment messages support multi-device push notifications.
Application scenarios:
The value of business reports lies not only in presenting insights and conclusions from data analysis, but also in driving a complete closed loop "from insight to decision-making to action".
-
Nomaly Inquiry: When reviewing operational reports, decision-makers need to be able to quickly raise questions with relevant personnel regarding unusual data fluctuations, requesting explanations and action plans.
-
Online Annotation: When reviewing analysis reports, if management has doubts about the metrics or analytical logic of a particular conclusion, they need to mark it directly in the original text and notify the analyst to provide supplementary explanations.
-
Anchored Discussion: The operations team holds multi-person discussions around the major promotion review report. All discussion content is linked to specific data charts, and can be fully traced afterward if needed.
Feature Overview:
Reports support comments
-
Comments Enabled: Report editors can manually enable the comments switch on the editing page. Once enabled, report viewers can post comments.
-
Text Comments: Hovering the mouse over the right side of a paragraph will bring up a comment entry point. You can also select part of the text to post a comment.
-
Chart Comments: Hovering the mouse over the right side of the chart will bring up a comment entry point, or you can submit a comment through the menu in the upper right corner of the chart.
-
Mixed Area Comments: Select a mixed area containing text and charts to initiate comments uniformly.
Supports mentioning users and providing instant notifications via DingTalk, Lark, and other channels.
-
@Users within the organization: Enter "@" in the comment input box to bring up a list of users within the organization (up to 50 users can be displayed), which supports keyword search.
-
Automatic push notifications: Sending notifications is selected by default when mentioning someone. When a comment is posted, the mentioned user will be simultaneously notified via third-party channels linked to the organization (DingTalk, Lark, etc.) or a custom channel.
A global comment panel allows you to view and manage all comments in one place.
-
Global Panel Entry: Accessed via the comment icon in the upper right corner of the report preview page. Opening it displays a list of all comments for the current report.
-
Filtering capabilities: Supports viewing comments by category: "All Comments", "Related to Me", and "Report Global Comments".
-
Jump to the comment entry: Clicking on a comment entry will automatically jump to the corresponding position in the main text on the left.
-
Initiate global comments: Comments can be initiated on the entire report in the editing box at the bottom of the panel.
4. Intelligent Text Processing: Five AI Functions for Intelligent Processing of Massive Text Volumes
Businesses can quickly transform large amounts of text data scattered in spreadsheets into structured business information that can be categorized, extracted, translated, and summarized, reducing manual processing costs and improving data flow efficiency and the quality of business decisions.
Feature 1: Five AI functions make text analysis accessible to everyone.
Application scenarios:
Business professionals often face the need to process large amounts of text data when using spreadsheets—such as categorizing and annotating customer feedback, extracting order information, translating multiple languages, and summarizing long texts. Traditional methods rely on complex formulas or even manual processing line by line, which is inefficient and prone to errors.
Function Overview:
Five AI Functions: Ushering in a New Paradigm for Intelligent Text Processing
-
Categorization and Tagging: Automatically recognizes text semantics and assigns preset tags, such as categorizing customer satisfaction (satisfied/neutral/unsatisfied) in the e-commerce industry. Supports custom categorization rules and batch tag editing options.
-
Information Extraction: Accurately extract key information from unstructured text, such as order number, customer name, contact information, etc.
-
Intelligent Translation: Supports text conversion between English, Simplified Chinese, Traditional Chinese, and Japanese, and allows customization of translation style, wording requirements, and length limits.
-
Content Summary: Intelligently summarizes long texts, extracts core points, and supports custom summary styles and focus areas.
-
Custom AI Functions: Customize your own AI tasks using natural language commands, support defining and referencing data variables, and meet personalized analysis needs.
Intelligent configuration; generated results can be modified and regenerated.
-
Preview and Verification: All AI functions support previewing the processing result of the first piece of data before execution, and can be applied in batches only after confirming that it meets expectations.
-
Flexible output: The result position can be selected as "to the right of the reference column" or "custom position" to adapt to different table layout needs.
-
Automatic Updates: When enabled, the AI processing results are updated in real time when the source data changes, eliminating the need for repeated operations.
-
Fine-grained management: Supports various management operations such as regenerating, editing configurations, deleting entire columns of results, and updating selected cells individually.
5. Open Ecosystem: Enabling Enterprise-Grade Data Analytics Capabilities for AI Agents Across the Platform
Feature 1: Quick BI Data Analysis Skills enable you to run a complete enterprise data analysis suite.
Pain points:
Using AI agents for data analysis is becoming increasingly common, but in practice, companies often encounter a series of real-world problems:
- Letting large models directly calculate metrics? This carries a high risk of misleading results and is unreliable.
- Feeding sensitive corporate data to generic agent tools? This will fail to meet security and compliance requirements.
- Faced with complex professional needs such as year-on-year and month-on-month comparisons and attribution analysis? General Prompt simply cannot handle it.
Application scenarios:
Embedding Quick BI intelligent analysis skills into the AI Agent workflow can expand the depth and breadth of data analysis performed by general AI agents, such as:
-
Monthly Business Review: How many goals were achieved, where did we fall short, and why? One round of discussion completes the analysis of the past half-day.
-
Multi-source data reporting: Excel targets + BI data + PDF invoices, automatically integrated to generate a monthly report that can be presented directly.
-
Scheduled automatic monitoring: Set up monthly automatic analysis, proactively alert you when achievement rates are abnormal, and analyze your own operation.
Function Overview:
-
Quick BI intelligent analysis skill supports the ability to call Smart Q Questions, Smart Q Reports, Smart Q Interpretations, and Smart Q Reports, and has automatic intent recognition function.
-
Questioning ：Ask questions about data using natural language; the system automatically understands intent and intelligently selects the appropriate data tables. It supports local Excel/CSV files and Quick BI enterprise datasets—requiring no SQL or data modeling—and delivers instant answers.
-
Interpretation：Automatically analyzes data files to uncover key changes, anomalous fluctuations, and trends, generating analytical conclusions accompanied by actionable recommendations. Eliminate the need to manually scan rows of data; anomalies surface automatically.
-
Dashboard：Uses natural language to deeply interpret the analytical logic and metric definitions within Quick BI dashboards. It automatically breaks down complex questions into multi-step analyses and identifies root causes, transforming static dashboards into dynamic, interactive Q&A interfaces.
-
Report：Generate professional analysis reports from multi-source data with a single click. Supporting imports in formats such as Excel, PDF, and Word, the system uses AI to integrate and cross-analyze data, producing well-structured, visually rich reports ready for use in executive briefings and decision-making.
Installation method:
- Skill Marketplace Launch: The "Smart Q" Data Analysis Agent is now available on QoderWork and the Wukong Skill Marketplace, supporting one-click download.
- Manual installation: Download the Skill file, extract it to a local folder, and manually upload it to the Skill Center.
Conclusion
The core of the V6.2 upgrade drives Quick BI’s intelligent assistant, "Smart Q," toward becoming a "trustworthy and operational enterprise-grade AI data analysis system."This entails three key shifts: enabling AI to truly comprehend an enterprise's specific business language rather than merely matching surface-level, general semantics; transforming analytics from reactive responses to proactive insights that automatically pinpoint the root causes of anomalies; and integrating AI and BI capabilities deeply into existing enterprise collaboration workflows and tool ecosystems, rather than operating them in isolation.
Ultimately, the issue that V6.2 aims to resolve is not whether AI is capable of performing analysis, but whether enterprises have the confidence to entrust analysis to AI.
To learn more about specific capabilities or request a product demo, please scan the QR code to connect with your dedicated account manager, or apply for a trial and solution consultation via the official Quick BI website.
Alibaba Cloud Model Studio
Big Data Consulting for Data Technology Solution
Big Data Consulting Services for Retail Solution
AI Acceleration Solution
