Docs

Intelligent Data Insight

Quick Start

Getting Started Guide for Business Users

Introduction

tip

DataWinds Intelligence) platform. Simply put, it allows you to turn your data into charts by simply dragging and dropping, and these charts can be grouped together to form a dashboard.

First, let us introduce the main modules of Smart Data Insight and their relationships. Understanding how Smart Data Insight transforms data into charts will help you better understand this product.
The entire workflow of Smart Data Insight is illustrated in the diagram below, which mainly involves five concepts:

Concept	Note
`Project`: There are many business lines in a company, and the data and dashboards for different business lines cannot be combined. Therefore, it is necessary to distinguish and isolate them. This is the purpose of projects. Each business line can create its own project, which facilitates resource management and access control.	Generally, you will not be involved in project management, so you do not need to be concerned about it. The only point to note is that resources in different projects are isolated. If you ever open DataWind and cannot find the dashboard you previously created, you should first check whether you have entered the wrong project.
`Data source`: To create a chart, you need data in the chart. The origin of this data is the data source. Commonly used tables, such as Lark tables, Hive tables, and ClickHouse tables, are all data sources.
`Dataset`: Since there are many types of data sources, some cannot be queried directly, and some are slow to query if connected directly. Therefore, DataWind needs a medium that can import data from data sources and store it for subsequent querying. This medium is called a dataset.	Tables you are most familiar with, such as Lark tables and Excel, can be made into datasets. To upload your own data, refer to this document: Dataset Overview.
`Visual query`: Once data is available in the dataset, you can query this data. This is visual query, which is where you query data and create charts.
`Dashboard`: After creating charts, you need a place to display them. This is the dashboard, which is essentially a blank canvas where you can arrange different charts, text, images, and more. This is what is commonly referred to as a dashboard.

The concepts above are primarily to help you understand what Smart Data Insight does. For most people, it is sufficient to be familiar with visual queries and dashboards, because in many enterprises, datasets are typically created by data warehouse staff in the big data team, so you can use them directly.
Next, we will guide you from importing data, through visual queries (the module we use most frequently), to dashboards, helping you start using Smart Data Insight from scratch. Let's get started!

1. Step 1: Import data

tip

To query data using DataWind, you first need data. Data is typically stored in data sources, so you must import data from the data source into DataWind as a dataset before you can perform queries.

For most users, the data warehouse team will help create datasets, and you can use them directly. If you do not usually need to create datasets, you can skip this section. To create your own dataset for use, refer to the following process.

1.1 Supported data sources for Smart Data Insight

Source name	Supported connection types	Supported environments	Creation supported Custom SQL	Integration help document
Excel/CSV file	Offline extraction, direct connection	SaaS, on-premises deployment	Not supported	Local Excel/CSV Files
Access	Offline extraction	On-premises deployment	Not supported	Access
Kafka	Real-time extraction	On-premises deployment	Not supported	Kafka
Pulsar	Real-time extraction	On-premises deployment	Not supported	Pulsar
GreenPlum	Offline extraction	SaaS, on-premises deployment	Supported	GreenPlum
SAP HANA	Offline extraction	SaaS, on-premises deployment	Supported	SAP HANA
Impala	Offline extraction, direct connection	SaaS, on-premises deployment	Supported	Impala
ClickHouse Community Edition	Offline extraction, direct connection	SaaS, on-premises deployment	Supported	ClickHouse
HBase Phoenix	Offline extraction	SaaS, on-premises deployment	Supported	Hbase Phoenix
HBase	Offline extraction	SaaS, on-premises deployment	Not supported	HBase
Vertica	Offline extraction	SaaS, on-premises deployment	Not supported	Vertica
Amazon Athena	Offline extraction	SaaS, on-premises deployment	Supported	Amazon Athena
Amazon Redshift	Offline extraction	SaaS, on-premises deployment	Supported	Amazon Redshift
ByteHouse CDW	Offline extraction, direct connection	SaaS, on-premises deployment	Supported	ByteHouse CDW
ByteHouse CE	Offline extraction, direct connection	SaaS, on-premises deployment	Supported	ByteHouse CE
Presto	Offline extraction, direct connection (supported from version V2.50.0 and later)	SaaS, on-premises deployment	Supported	Presto
Databricks	Offline extraction	SaaS, on-premises deployment	Supported	Databricks
AnalyticDB for MySQL 2.0	Offline extraction	SaaS, on-premises deployment	Supported	AnalyticDB for MySQL 2.0
AnalyticDB for MySQL 3.0	Offline extraction	SaaS, on-premises deployment	Supported	AnalyticDB for MySQL 3.0
Hologres	Offline extraction, direct connection	SaaS, on-premises deployment	Supported	Hologres
MySQL	Offline extraction, direct connection	SaaS, on-premises deployment	Supported	MySQL
IBM DB2	Offline extraction	SaaS, on-premises deployment	Supported	IBM DB2
Oracle	Offline extraction	SaaS, on-premises deployment	Supported	Oracle
PostgreSQL	Offline extraction, direct connection	SaaS, on-premises deployment	Supported	PostgreSQL
OpenGauss	Offline extraction	SaaS, on-premises deployment	Supported	OpenGauss
SQL Server	Offline extraction	SaaS, on-premises deployment	Supported	SQL Server
Teradata	Offline extraction	SaaS, on-premises deployment	Supported	Teradata
Doris	Offline extraction, direct connection	SaaS, on-premises deployment	Supported	Doris
Damon	Offline extraction	SaaS, on-premises deployment	Supported	Damon
TiDB	Offline extraction	SaaS, on-premises deployment	Supported	TiDB
GoldenDB	Offline extraction	SaaS, on-premises deployment	Supported	GoldenDB
OceanBase MySQL	Offline extraction	SaaS, on-premises deployment	Supported	OceanBase MySQL
OceanBase Oracle	Offline extraction	SaaS, on-premises deployment	Supported	OceanBase Oracle
Hive	Offline extraction	SaaS, on-premises deployment	Supported	Hive
MaxCompute	Offline extraction	SaaS, on-premises deployment	Supported	MaxCompute
Transwarp Inceptor	Offline extraction	SaaS, on-premises deployment	--	Transwarp Inceptor
StarRocks	Offline extraction, direct connection	SaaS, on-premises deployment	Not supported	StarRocks
Huawei FusionInsight	Offline extraction	SaaS, on-premises deployment	Supported	Huawei FusionInsight
Tencent Cloud DLC	Offline extraction	SaaS, on-premises deployment	Supported	Tencent Cloud DLC
DataFinder	Direct connection	SaaS, on-premises deployment	Not supported	DataFinder Data Integration
LAS (public cloud version)	Offline extraction	SaaS, on-premises deployment	Supported	LAS
Intelligent outbound call	Offline extraction	SaaS, on-premises deployment	--
Metadata connection and application	--	On-premises deployment	--	Metadata Connection and Application
Douyin ecosystem data	Offline extraction	SaaS, on-premises deployment	Not supported	None
Ocean Engine/Ocean Qianchuan	Offline extraction	SaaS, on-premises deployment	Not supported	Ocean Engine/Ocean Qianchuan
REST API	Offline extraction	SaaS, on-premises deployment	Not supported	REST API 2.0
Content platform	Offline extraction	SaaS, on-premises deployment	Not supported
MongoDB	Offline extraction	SaaS, on-premises deployment	Supported	MongoDB
WeChat Official Account	Offline extraction	SaaS, on-premises deployment	Not supported	WeChat Official Account
Constellation data	Direct connection	SaaS, on-premises deployment	Not supported	Constellation data
Calendar data	Direct connection	SaaS, on-premises deployment	Not supported	Calendar data
Lark spreadsheet	Offline extraction	SaaS, on-premises deployment	Not supported	Lark Sheets
Lark multidimensional table	Offline extraction	SaaS, on-premises deployment	Not supported	Lark Base Ta
Lark Project (Meego) data	Offline extraction	SaaS, private deployment	Not supported	Lark Project (Meego) data

1.2 Configuration process

For specific configuration, refer to the help documentation provided after the supported data sources above.

2. Step 2: Using visual queries

2.1 Visual query page layout

tip

Now we have arrived at the most important module for everyone: visual query, which is where we query data.

First, the overall layout of the visual query page: the visual query page is divided into four areas:
A. Dataset information and field list: Displays which dataset you are querying, which fields are available in this dataset, and allows you to switch datasets here.
B. Chart and analysis features: Select chart types, configure charts, and choose analysis features (such as sorting, percentage, year-over-year and period-over-period comparison, table calculations, and more).
C. Chart display area: Place the fields you want to query here. This area shows the final chart style and allows you to adjust the chart's data format.
D. Toolbar: Query history, download data, configure chart data monitoring, and save charts.

2.2 Concepts of dimensions and metrics

tip

Before performing a query, there is a small concept that everyone needs to learn, which is the concept of dimensions and metrics. Dimensions and metrics are generally collectively referred to as fields.

Refer to the two figures below. On the left, I have placed all the fields from a dataset in DataWind into a chart, while on the right is the underlying data of this dataset. You will notice that each field here corresponds to a specific column in the original data. Moreover, when the chart type selected in DataWind is a standard table, the queried data is very similar to the original data (although they look the same, there are actually differences, which we will discuss later).
Therefore, when selecting fields in DataWind, it is essentially equivalent to querying specific columns from the underlying data:

Next is the difference between dimensions and metrics, which is quite straightforward. You can think of a dimension as the level at which you view the data, and a metric as a measure. For example, if you want to see the grades of all students in a class, the student name is the dimension, and the grade is the metric. Dimensions and metrics are not fixed; sometimes a dimension field can also serve as a metric. For instance, if you want to see how many students fall into each grade range, the grade range is the dimension, and the count of student names is the metric.
The same applies in DataWind: even if a field is classified as a dimension, it can still be placed in the metrics section, and metric fields can also be placed in the dimensions section.

2.3 Steps for querying data

tip

Next are the steps for using visual query. Here, we will use a hypothetical requirement as an example to walk you through how to use DataWind to complete a data query.

2.3.1 Process for completing a query

Query example 1

Suppose our current data query requirement is: In Q2 2024, for accounts with more than 1,000 followers, observe the trend of review volume changes across all secondary verticals under the music primary video vertical. Upon receiving this requirement, it can first be broken down as follows: Filter conditions: Q2 2024, followers greater than 1,000, primary video vertical is music Dimensions: secondary vertical, date (although not stated in the requirement, date is necessary to observe trends) Metric: review volume Objective: observe the trend
Step 1: First, place the required dimensions and metrics in the dimensions and metrics section. You can either drag them or click the fields on the left.
Step 2: Then configure the filters. You can also drag the fields to be filtered into the filter section and select the appropriate filtering method.
Step 3: Finally, select a suitable chart. Since the goal is to observe changes in trends, a line chart was selected.

Suppose our current data query requirement is: In Q2 2024, for accounts with more than 1,000 followers, observe the trend of review volume changes across all secondary verticals under the music primary video vertical.

Upon receiving this requirement, it can first be broken down as follows:
- Filter conditions: Q2 2024, followers greater than 1,000, primary video vertical is music
- Dimensions: secondary vertical, date (although not stated in the requirement, date is necessary to observe trends)
- Metric: review volume
- Objective: observe the trend

Step 1: First, place the required dimensions and metrics in the dimensions and metrics section. You can either drag them or click the fields on the left.

Step 2: Then configure the filters. You can also drag the fields to be filtered into the filter section and select the appropriate filtering method.

Step 3: Finally, select a suitable chart.

Since the goal is to observe changes in trends, a line chart was selected.

Query example 2

Suppose our current data query requirement is: Based on the granularity of the primary video vertical, check the proportion of videos sourced from Douyin this year. Upon receiving this requirement, it can first be broken down as follows Filter conditions: this year, video source is Douyin Dimension: primary video vertical Metric: video count Objective: observe the proportion
First, place the required dimensions and metrics into the dimensions and metrics panel. You will notice that this dataset does not have a metric for the number of videos, but it does have a video ID field. By counting the video IDs, you obtain the number of videos. Therefore, you can drag this field into the metrics panel and then select count.
You can configure filters.
Finally, select an appropriate chart. Since you want to view proportions, you can choose a pie chart.

Suppose our current data query requirement is: Based on the granularity of the primary video vertical, check the proportion of videos sourced from Douyin this year.

Upon receiving this requirement, it can first be broken down as follows
- Filter conditions: this year, video source is Douyin
- Dimension: primary video vertical
- Metric: video count
- Objective: observe the proportion

First, place the required dimensions and metrics into the dimensions and metrics panel.

You will notice that this dataset does not have a metric for the number of videos, but it does have a video ID field. By counting the video IDs, you obtain the number of videos. Therefore, you can drag this field into the metrics panel and then select count.

You can configure filters.

Finally, select an appropriate chart.

Since you want to view proportions, you can choose a pie chart.

2.3.2 Three concepts to learn

How to distinguish field types	The fields in Smart Data Insight are categorized just like in Excel spreadsheets. They can be distinguished by the icon in front of the field name, as shown in the figure below: Red indicates date, green indicates text, and yellow indicates numeric values (including both decimals and integers). The main thing to note is that different field types require different filtering methods. For example, date fields can be filtered by the most recent N days. However, if a field is named 'date' but is actually of text type, you can only select individual values when filtering. To enable date filtering, contact the dataset owner to make adjustments.
What does sum above the metric mean?	If you are new to Smart Data Insight, you may be confused about why 'sum' appears above a field in the metrics panel, and sometimes 'count' or 'aggregate' is shown. Here, there is a concept called aggregation, which is actually easy to understand. When a set of data is calculated to produce a single value, this calculation process is called aggregation. For example, the common sum operation calculates a single value from a set of numbers; finding the maximum value also produces a single value from a set; counting also results in a single value from a set. This is what aggregation means. Since users rarely view detailed data in their daily work, DataWind will automatically aggregate the metrics you add to any chart other than the detail table, based on the dimensions you select. For example: On the left below is a query in DataWind. Here, the dimension is primary business, and the metric is review volume. The review volume is automatically summed, so what you see is the total review volume under each primary business. The right side displays the underlying raw data, which consists of a large number of records. Of course, users may not always need summed data. You can click the downward triangle in front of the metric to select the aggregation method; for example, to view quantity, select count; to view the average, select average. As mentioned earlier, fields have different types: Numeric types allow mathematical operations such as sum. Text types can only be used for operations like count. There is a special case: if the expression for a field already uses an aggregation function (such as sum or count), SQL syntax does not allow nested aggregation. Therefore, it will automatically display "Aggregation," and you cannot select other aggregation methods. Side note: Fields with aggregation functions in their expressions cannot be placed in the dimension column!
What is the p_date filter	Attentive users may notice that the filter bar in DataWind visual queries always automatically includes a filter called p_date. This filter cannot be removed from the filter list, meaning it must be applied. Users often wonder what p_date is when using the product. In databases, there is a concept called partition. You can think of it as a sheet in a Lark table. Each day, a new sheet is added, and the data for that day is placed inside it. To make it easier to distinguish between these partitions, each partition is named after the date of that day. Here, p_date actually refers to the partition, meaning the date when the data was updated. Filtering by a specific p_date means viewing the data updated on that particular day. There is also a related concept here: incremental datasets and full datasets. Incremental table: Each partition only contains the data for the current day. Full table: Each partition contains all data from historical records up to the present. When querying a full dataset, it is sufficient to filter the partition by the most recent day with data. When querying an incremental dataset, the partition filter should cover the entire period you want to query.

How to distinguish field types

The fields in Smart Data Insight are categorized just like in Excel spreadsheets. They can be distinguished by the icon in front of the field name, as shown in the figure below:

Red indicates date, green indicates text, and yellow indicates numeric values (including both decimals and integers).
The main thing to note is that different field types require different filtering methods.
- For example, date fields can be filtered by the most recent N days. However, if a field is named 'date' but is actually of text type, you can only select individual values when filtering. To enable date filtering, contact the dataset owner to make adjustments.

What does sum above the metric mean?

If you are new to Smart Data Insight, you may be confused about why 'sum' appears above a field in the metrics panel, and sometimes 'count' or 'aggregate' is shown.

Here, there is a concept called aggregation, which is actually easy to understand. When a set of data is calculated to produce a single value, this calculation process is called aggregation.

For example, the common sum operation calculates a single value from a set of numbers; finding the maximum value also produces a single value from a set; counting also results in a single value from a set. This is what aggregation means.

Since users rarely view detailed data in their daily work, DataWind will automatically aggregate the metrics you add to any chart other than the detail table, based on the dimensions you select.
For example:

On the left below is a query in DataWind. Here, the dimension is primary business, and the metric is review volume. The review volume is automatically summed, so what you see is the total review volume under each primary business.
The right side displays the underlying raw data, which consists of a large number of records.

Of course, users may not always need summed data. You can click the downward triangle in front of the metric to select the aggregation method; for example, to view quantity, select count; to view the average, select average.
As mentioned earlier, fields have different types:

Numeric types allow mathematical operations such as sum.
Text types can only be used for operations like count.
There is a special case: if the expression for a field already uses an aggregation function (such as sum or count), SQL syntax does not allow nested aggregation. Therefore, it will automatically display "Aggregation," and you cannot select other aggregation methods.
- Side note: Fields with aggregation functions in their expressions cannot be placed in the dimension column!

What is the p_date filter

Attentive users may notice that the filter bar in DataWind visual queries always automatically includes a filter called p_date. This filter cannot be removed from the filter list, meaning it must be applied.

Users often wonder what p_date is when using the product.
In databases, there is a concept called partition. You can think of it as a sheet in a Lark table. Each day, a new sheet is added, and the data for that day is placed inside it. To make it easier to distinguish between these partitions, each partition is named after the date of that day. Here, p_date actually refers to the partition, meaning the date when the data was updated. Filtering by a specific p_date means viewing the data updated on that particular day.
There is also a related concept here: incremental datasets and full datasets.

Incremental table: Each partition only contains the data for the current day.
Full table: Each partition contains all data from historical records up to the present.

When querying a full dataset, it is sufficient to filter the partition by the most recent day with data. When querying an incremental dataset, the partition filter should cover the entire period you want to query.

tip

After completing the query, you can choose to save the current chart, download the data, or share the query results with other users.

When saving, if you have already created a dashboard, you can directly select this dashboard to save to it.
When downloading, you can choose to download the data. If it is a chart, you can also choose to download the image.
Query results in Smart Data Insight can be shared via a browser link.

Save chart After the chart is created, you can click Save on the toolbar to save it as a chart. If a dashboard has already been created, you can directly select the dashboard to quickly add the chart to it. If a dashboard has not been created or you do not intend to display it on a dashboard, you do not need to select one. Later, you can find the chart in Chart Management.
Download data After creating a chart, if you do not need to save the chart, you can also choose to download the data or download a screenshot of the chart. Caution: When downloading a pivot table, make sure to select the pivot table format. Otherwise, the downloaded file will not be a pivot table.
Share query In DataWind, each time you click to run a query, a unique query link is generated. You can directly copy the browser link, and when other users open it, they will enter the same query page as you. If you obtain another user's query link, you can continue querying based on their results. If the top displays an unsaved query, you can freely modify and save it without affecting other users. If the top displays a chart name and the Save As button in the upper right corner is clickable, you can freely modify the query conditions, just do not click the Save button, otherwise you will overwrite the saved conditions of others. Of course, clicking Save As to save as your own chart is fine.

Save chart

After the chart is created, you can click Save on the toolbar to save it as a chart.
If a dashboard has already been created, you can directly select the dashboard to quickly add the chart to it.
If a dashboard has not been created or you do not intend to display it on a dashboard, you do not need to select one. Later, you can find the chart in Chart Management.

Download data

After creating a chart, if you do not need to save the chart, you can also choose to download the data or download a screenshot of the chart.
Caution: When downloading a pivot table, make sure to select the pivot table format. Otherwise, the downloaded file will not be a pivot table.

Share query

In DataWind, each time you click to run a query, a unique query link is generated. You can directly copy the browser link, and when other users open it, they will enter the same query page as you.
If you obtain another user's query link, you can continue querying based on their results.
- If the top displays an unsaved query, you can freely modify and save it without affecting other users.
- If the top displays a chart name and the Save As button in the upper right corner is clickable, you can freely modify the query conditions, just do not click the Save button, otherwise you will overwrite the saved conditions of others. Of course, clicking Save As to save as your own chart is fine.

2.5 Common chart configurations

tip

Here are some commonly used chart configuration options. You can click the blue text to jump to detailed help documentation.

Configuration	Explanation
Change 2.2 Modify dimensions and measures You can adjust the unit, format, prefix, suffix, decimal places, and more for data in the chart.
Change Dimension Field Configuration You can quickly modify the date granularity of date-type fields in the chart.
Change Configure field information You can modify the header names in the chart. This is only a temporary change and does not affect the actual field names.
Quick sorting Whether in visual queries or dashboards, move the mouse to the right side of the corresponding header and click the icon that appears to trigger the quick sort popup. Here, you can quickly sort the data.
Display data on the chart This allows each data point in the chart to display its specific value.

2.6 Advanced features for visual queries

tip

Here, we mainly introduce some advanced features of the visualization query page. These features are not necessarily required, but if you encounter similar scenarios, they can be helpful for you~

2.6.1 Chart configuration

Configuration	Description
Dynamic Dimensions and Dynamic Metrics To replace the dimensions or metrics in a chart without modifying the chart itself, try this feature. A common scenario is placing dates with different time granularities in dynamic dimensions, allowing you to quickly switch date granularity on the dashboard. Example link: Dynamic Dimensions and Dynamic Metrics
Drill Up and Drill Down When performing analysis, for example, if you notice abnormal data in a primary business line and want to examine the secondary business lines under it in detail, you can use this feature. Analyze other dimensions by drilling down without modifying the chart.
Overview of Field Configuration This feature is relatively simple. It makes the data in the chart easier to distinguish and improves the appearance of the chart. Currently, it supports adding single color, icon sets, color scales, and data bars to charts. Example link: Configure field information

Configuration

Description

Dynamic Dimensions and Dynamic Metrics

To replace the dimensions or metrics in a chart without modifying the chart itself, try this feature.
A common scenario is placing dates with different time granularities in dynamic dimensions, allowing you to quickly switch date granularity on the dashboard.
Example link: Dynamic Dimensions and Dynamic Metrics

Drill Up and Drill Down

When performing analysis, for example, if you notice abnormal data in a primary business line and want to examine the secondary business lines under it in detail, you can use this feature.
Analyze other dimensions by drilling down without modifying the chart.

Overview of Field Configuration

This feature is relatively simple. It makes the data in the chart easier to distinguish and improves the appearance of the chart. Currently, it supports adding single color, icon sets, color scales, and data bars to charts.
Example link: Configure field information

2.6.2 Filter configuration

Configuration	Description
Place filters on the chart By selecting "Show as chart filter", you can place the filter box on the chart, making it easier to quickly switch filters.
Combination Filter In the filter bar of DataWind, multiple filters are set to an 'and' relationship by default. To achieve more complex filtering, use the composite filtering feature.
Subfilters Sub-filters enable cascading filters. For example, if you set province and city filters to cascade, after selecting a province, only cities within that province are displayed in the city filter.

Configuration

Description

Place filters on the chart

By selecting "Show as chart filter", you can place the filter box on the chart, making it easier to quickly switch filters.

Combination Filter

In the filter bar of DataWind, multiple filters are set to an 'and' relationship by default. To achieve more complex filtering, use the composite filtering feature.

Subfilters

Sub-filters enable cascading filters. For example, if you set province and city filters to cascade, after selecting a province, only cities within that province are displayed in the city filter.

2.6.3 Analysis feature

Sorting, percentage, period-over-period comparison, table calculation, comparison, and more

Previously, DataWind provided dedicated video tutorials for each analysis feature.

2.6.4 Chart monitoring

Monitor data in charts and trigger alerts

DataWind supports monitoring and alerting. This feature (dimensions must include a date field) supports configuring metric fluctuation monitoring and dimension enumeration value change monitoring. For example, when a metric fluctuation reaches a certain threshold, users or groups will automatically receive alert notifications.

3. Step 3: Using dashboards

3.1 Create a blank dashboard

tip

This chapter introduces the dashboard feature of DataWind. Previously, we learned how to use visual queries to obtain data, and the charts created can be saved. These charts need a unified place for display—that is, the dashboard. The dashboard is essentially a blank canvas where you can arrange components such as charts, text, and images. For tips and procedures on building dashboards, you can also refer to Dashboard Overview.

3.2 Personal dashboards and public dashboards

tip

Since creating a dashboard is equivalent to adding a new resource under a project, project edit permissions are required. However, some users may only want to create dashboards for personal use without requesting additional permissions. Therefore, DataWind provides both personal dashboards and public dashboards. Note the following:
A personal dashboard can grant others view permissions at most. To allow other users to collaboratively edit and maintain the dashboard, a public dashboard must be created.

	Personal dashboard	Public dashboard
Permissions required for creation	Project view permission	Project edit permission
Maximum permission that can be granted to others	View permission	Manage permission

As for how to choose between creating a personal or a public dashboard, you will be prompted to select the dashboard's save path during creation. If you choose the personal dashboard path, the dashboard created will be a personal dashboard; otherwise, it will be a public dashboard.

If you later move a dashboard from the personal dashboard path to the public dashboard path, it will become a public dashboard, and vice versa.

3.3 Edit dashboards

tip

After creating a blank dashboard, you can start editing the dashboard. Editing a dashboard is actually quite simple. If you have ever created a PowerPoint presentation, adding elements here is just like adding things in PowerPoint.

Step 1: Enter the editing page

Click Edit in the upper right corner of the blank dashboard to enter the editing page.
First, get familiar with the layout of the page. The left side is the component panel, which shows what you can add to the dashboard, such as charts, images, text, web pages, and more. The bottom area is where you add Sheet pages. For example, to divide a dashboard into multiple pages, you can use this area. The top area is the toolbar, where you can adjust layers and similar settings. The right side is for page configuration, such as layout style, background color, and so on.

Click Edit in the upper right corner of the blank dashboard to enter the editing page.

First, get familiar with the layout of the page.

The left side is the component panel, which shows what you can add to the dashboard, such as charts, images, text, web pages, and more.
The bottom area is where you add Sheet pages. For example, to divide a dashboard into multiple pages, you can use this area.
The top area is the toolbar, where you can adjust layers and similar settings.
The right side is for page configuration, such as layout style, background color, and so on.

Step 2: Add charts and other components

(1) Click Add Chart on the left, or drag and drop it onto the canvas in the center to add a chart component to the canvas.
(2) After adding a chart component, you can configure the corresponding chart for this component. There are two ways to do this.	The first way is to search for charts you have previously saved on the right and add them directly. The second way is to click Add Chart in the middle of the component to enter the query page and create a chart on the spot.
(3) After adding charts to the dashboard, you can also add other components, such as titles, images, or notes, to provide explanations or make the structure clearer if needed.
(4) When you finish creating the dashboard, click Save in the upper right corner.

(1) Click Add Chart on the left, or drag and drop it onto the canvas in the center to add a chart component to the canvas.

(2) After adding a chart component, you can configure the corresponding chart for this component. There are two ways to do this.

The first way is to search for charts you have previously saved on the right and add them directly.

The second way is to click Add Chart in the middle of the component to enter the query page and create a chart on the spot.

(3) After adding charts to the dashboard, you can also add other components, such as titles, images, or notes, to provide explanations or make the structure clearer if needed.

(4) When you finish creating the dashboard, click Save in the upper right corner.

Step 3: Configure some minor dashboard settings

Add Sheet pages You can add multiple Sheet pages to a dashboard and place different charts on each page. This is useful for scenarios where you want to display multiple categories of data in a single core dashboard.
Dashboard Theme To make the dashboard more visually appealing, you can use the theme feature of Smart Data Insight to configure the color scheme, fonts, data formats, and more for all charts with one click.
Layout style Smart Data Insights supports two layout modes Tile layout: Dashboard components automatically adjust their layout when dragged, making it easy to quickly arrange them Free layout: As the name suggests, all components can be freely dragged to any position, making it suitable for scenarios with special layout requirements
Global Filter Sometimes, a dashboard may contain many charts, and when viewing data, users may want to change the filter conditions for all charts with a single click. This can be achieved using the global filter

Add Sheet pages

You can add multiple Sheet pages to a dashboard and place different charts on each page. This is useful for scenarios where you want to display multiple categories of data in a single core dashboard.

Dashboard Theme

To make the dashboard more visually appealing, you can use the theme feature of Smart Data Insight to configure the color scheme, fonts, data formats, and more for all charts with one click.

Layout style

Smart Data Insights supports two layout modes
- Tile layout: Dashboard components automatically adjust their layout when dragged, making it easy to quickly arrange them
- Free layout: As the name suggests, all components can be freely dragged to any position, making it suitable for scenarios with special layout requirements

Global Filter

Sometimes, a dashboard may contain many charts, and when viewing data, users may want to change the filter conditions for all charts with a single click. This can be achieved using the global filter

3.4 Common uses of dashboards

In addition to providing a unified view of chart data, dashboards also have several other uses:

Copy dashboards or charts In many scenarios, dashboards or charts can be reused. For example, if someone else's chart is well made and you want to reuse it, Smart Data Insights also provides several copy features These copy features have certain permission requirements, such as for datasets or dashboards. If you find that this option is not available when you click on it, it means you do not have permission	The first method allows you to copy a chart from the current dashboard to another dashboard (the copied chart is a duplicate) The second method allows you to directly copy the entire dashboard (you can choose whether to create chart duplicates; if duplicates are not created, modifying a chart in one dashboard will synchronize the corresponding chart in both dashboards). The third method allows you to copy a sheet page to another dashboard (you can choose whether to create chart duplicates; if duplicates are not created, modifying a chart in one dashboard will synchronize the corresponding chart in both dashboards)
Configure dashboard: Data Subscription The product supports dashboard subscription configuration. After configuration, the dashboard will automatically send its content as a screenshot every day via Lark, email, or other office software to a specific user or group chat, making it easier for everyone to view the data daily.
Embed the dashboard's Chart Embedding The charts placed on the dashboard can be embedded elsewhere. For example, when writing a weekly report, you can insert them into a Lark document and view the data just as you would on the dashboard.

Copy dashboards or charts

In many scenarios, dashboards or charts can be reused. For example, if someone else's chart is well made and you want to reuse it, Smart Data Insights also provides several copy features
These copy features have certain permission requirements, such as for datasets or dashboards. If you find that this option is not available when you click on it, it means you do not have permission

The first method allows you to copy a chart from the current dashboard to another dashboard (the copied chart is a duplicate)

The second method allows you to directly copy the entire dashboard (you can choose whether to create chart duplicates; if duplicates are not created, modifying a chart in one dashboard will synchronize the corresponding chart in both dashboards).

The third method allows you to copy a sheet page to another dashboard (you can choose whether to create chart duplicates; if duplicates are not created, modifying a chart in one dashboard will synchronize the corresponding chart in both dashboards)

Configure dashboard: Data Subscription

The product supports dashboard subscription configuration. After configuration, the dashboard will automatically send its content as a screenshot every day via Lark, email, or other office software to a specific user or group chat, making it easier for everyone to view the data daily.

Embed the dashboard's Chart Embedding

The charts placed on the dashboard can be embedded elsewhere. For example, when writing a weekly report, you can insert them into a Lark document and view the data just as you would on the dashboard.

4. Get familiar with the permission system

4.1 Permission description

There are two types of permissions in the product:
- The first type is resource permissions. For example, dashboards and datasets are both referred to as resources in DataWind. Viewing, editing, and managing these resources are all considered resource permissions.
- The second type is row and column permissions. Only datasets have row and column permissions, which determine which rows of data users can view in a dataset.
- Normally, if a user has dashboard permissions, they can view chart data even without dataset permissions for the charts on the dashboard. However, if the dataset is configured with row and column permissions, the user must also have the dataset's row and column permissions to view the data.

4.2 How to grant permissions to others

Caution: Only dashboard or dataset administrators can proactively grant permissions to others.
Normally, the creator of a resource becomes its owner and is granted management permissions. However, the owner does not always have management permissions. For example, if the project is under moderate or higher control, the resource creator will have at most editing permissions.

Dashboard
Dataset

4.3 How to request permissions

Project Center - Permission Request

5. How to seek help when encountering problems

tip

During use, users may encounter various issues. Since DataWind mainly serves as a platform for querying data and gaining business insights, the data available for queries may be maintained by business teams within the enterprise. Therefore, some issues encountered are related to product functionality, while others are business-related. How can users correctly seek help when encountering problems to improve efficiency?

DataWind product usage issues This includes issues such as not knowing how to use a feature, feature malfunctions, feature consultations, submitting requests, and other product-related problems.	DataWind provides a help documentation center, where you can search for keywords to find solutions: User Guide--DataWind. If the issue still cannot be resolved, you can consult after-sales support via the bottom right corner of the DataWind product page.
Business issues This includes situations where the dataset data does not match the online data, unclear field definitions, or not knowing what certain fields mean.	If such issues are encountered while viewing a dashboard, you can consult the dashboard owner. If such issues are encountered while performing a query, you can consult the dataset owner.

DataWind product usage issues

This includes issues such as not knowing how to use a feature, feature malfunctions, feature consultations, submitting requests, and other product-related problems.

DataWind provides a help documentation center, where you can search for keywords to find solutions: User Guide--DataWind.
If the issue still cannot be resolved, you can consult after-sales support via the bottom right corner of the DataWind product page.

Business issues

This includes situations where the dataset data does not match the online data, unclear field definitions, or not knowing what certain fields mean.

If such issues are encountered while viewing a dashboard, you can consult the dashboard owner.
If such issues are encountered while performing a query, you can consult the dataset owner.

Last updated: 2026.06.11 17:26:58

Intelligent Data Insight

Introduction #

1. Step 1: Import data #

1.1 Supported data sources for Smart Data Insight #

1.2 Configuration process #

2. Step 2: Using visual queries #

2.1 Visual query page layout #

2.2 Concepts of dimensions and metrics #

2.3 Steps for querying data #

2.3.1 Process for completing a query

2.3.2 Three concepts to learn

2.4 Saving, downloading, and sharing #

2.5 Common chart configurations #

2.6 Advanced features for visual queries #