这是用户在 2024-3-19 16:15 为 https://confluence.shopee.io/display/CCDP/%5BPRD%5D+Monitoring+and+Alerting+in+SSAD 保存的双语快照页面，由 沉浸式翻译 提供双语支持。了解如何保存？
Skip to end of metadata
Go to start of metadata

Target Release

2024 Q2

Epic

SPCSP-58810 - [Data Platform] Monitoring & Alerting in SSAD Waiting  

Status

DRAFT

BRD

[BRD] RTFM Spike Monitoring and Handling

Design Link

PD Design: 

PM Draft Figma

PM

BPM: Phoebe Heng 

DPM: Faith Eng 

PD
Mia Zhao 

Developers

FE:

BE:

QATBD
UAT Test CaseTBD

Version History

Version

Description

Created By

Updated Date

v1.0Monitoring and Alerting in SSAD – Phase 1 initial draft

 Faith Eng 

2024-02-15

Background

Self-Service Analysis Dashboard (SSAD) currently has a function known as "Action Box" which displays a box, listing recommended actions on a chart for the dashboard visitors to consider when identifying specific trends on the chart. However, as this feature was pushed from a top-down approach, there were several pain points faced by the user: 
自助分析仪表板 (SSAD) 目前具有称为“操作框”的功能，该功能会显示一个框，在图表上列出建议的操作，供仪表板访问者在识别图表上的特定趋势时考虑。然而，由于该功能是自上而下推送的，因此用户面临着几个痛点：

  1. Business team only provided five actions for locals to configure, but these options are outdated and irrelevant for some of RTFM's team use cases. 
    业务团队只提供了五个操作供当地人配置，但这些选项已经过时，并且与 RTFM 的一些团队用例无关。
  2. Lengthy 2-step process to reveal the action details, which may not be very intuitive for users.
  3. Many users may not be aware of this function as it is tucked away at the bottom of 'Customize Tab', leading to low adoption.

Objectives

In early Q2 2024, Alert in SSAD aims to achieve the following: 

  • Streamline monitoring, alerting and action handling process by linking conditional formatting and target line into 1 same function
  • Introduce new list of actions using the playbook as a guideline for locals to handle spikes
  • Support an alert function in SSAD to minimise the time between detection and response 
  • Increase the visibility of action handling and alert configuration 

Target Users 

High Level User Flow from BPM 

Role
Responsibility
RTFM

  1. Monitor incoming traffic, analyse the spike and identify any bug as fast as possible

  2. Spike Management: Take responsive actions upon detecting elevated spike levels or abnormalities
Other SSAD Users
  1. Highlight underperforming metrics or provide an acceptable range for a particular metric
  2. Provide data interpretation and make data-driven business decisions

To-Be Process 

Navigation 

Module
Description
UI Reference
SSADEdit a dashboard > Monitoring Tab 

Function List

ModuleFunction CodeFunction NameDescription
SSADF001Monitoring ConditionsSet up the targeted metric in a chart and its monitoring rules to set up Conditional Formatting or Target Line
F002Alert Setting
  1. Define check frequency and alert frequency 
  2. Configure the display of the alert message and its suggested actions if any
  3.  Send notification to a targeted Seatalk group chat 
F003View / Edit Configuration
  1. Once the above are completed, user can view the finalised configurations and edit them
  2. Support display/hide function to show or hide configured target lines / conditional formatting
F004TutorialIntroduce the new monitoring and alerting function via a three-step pop-up window
User Tracking DataF005Tracking Data and MetricsTMS for tracking data. Describe tracking requests to evaluate adoption rate of this feature

Function Description

F01 – Monitoring Conditions

Monitoring Conditions consists of the following components: 

  • F010 – Overview 
  • F011 – Monitoring Tab
  • F012 – Metric Monitoring
  • F013 – Existing Configurations 

Component
Description
UI Reference

F010 – Overview

[AS-IS] Each of the different chart types in SSAD supports either target line or conditional formatting function.

  1. When the chart type is switched from those with target line to those with conditional formatting (and vice versa), the previous configuration for its target line/conditional formatting is removed and user will have to re-configure it again for the new chart type
  2. Target line (Fig 1a) and conditional formatting (Fig 1b) can be set up when dashboard is in draft or published status

[TO-BE]

  1. Support a tab and sliding pane to allow user to set up target line / conditional formatting and alert setting (alert box and alert notification) (optional)
  2. To set up a target line or conditional formatting, users will follow the same step by configuring target metric and its thresholds. 
  3. Dynamic target line (Fig 1c) will be removed due to its low usage (refer to query below) and to better facilitate the new monitoring process.
    1. Query to retrieve dynamic target line usage Expand source
  4. When a user switches to another chart type, the configured metric thresholds should persist.

Fig 1a: AS-IS target line configuration

Fig 1b: AS-IS conditional formatting configuration

Fig 1c: AS-IS dynamic target line 

F011 – Monitoring Tab
  1. On top of Data tab and Customize tab, introduce a new tab called "Monitoring" / "Monitor" (TBC) (Fig 2)
  2. When there are no configurations, display supporting text: 

    1. Monitor a metric's thresholds and set up alerts to be notified. Start now!

  3. Support a "+ Add Monitoring" button
    1. A sliding pane "Add Monitoring" will appear for user to configure monitoring and alert setting

Fig 2: To begin, user has to click onto "+ Add Monitoring" button from the monitoring tab

F011 – Metric Monitoring 

Support user to set up Conditional Formatting or Target Line by defining the 1) Target Metric and 2) metric thresholds. 

Subcomponent
Description
Title
  1. Title: Monitoring Thresholds 
  2. Depending on the chart type, display a tag to indicate whether conditional formatting or target line will be displayed on the chart visualisation.  
    1. Conditional Formatting (Fig 3a, Green):
      1. Applicable to Metric Card, Table & Pivot Table
      2. [Hover Text] Displays conditional formatting on {current chart type}, emphasizing data trends on your chart visualisation through dynamic colour adjustments based on set metric thresholds.
    2. Target Line (Fig 3b, Green):
      1. Applicable to Bar, Row, Stacked Bar, Line Chart
      2. [Hover Text] Displays target line(s) on {current chart type}, providing a visual reference line in the chart to compare actual values against metric thresholds.  
    3. No Visualisation
      1. CF / TL will not be applicable to Pie Chart, however the alerts set can still be triggered. 
Target Metric
  1. Same as current SSAD V1.0 F0350 Conditional Formatting and F0351 Target Line 
  2. Support dropdown to display all available metrics from the chart
    1. Includes customised metrics 
    2. [Placeholder Text for CF] Select Metric (Fig 3a, Yellow)
    3. [Placeholder Text for Target Line] Select Metric / Y-Axis (Fig 3b, Yellow)
  3. Listed metric name should include the aggregation method
    1. e.g. Status (by Count)
  4. By default, if there is only 1 metric available, directly select that for user.
  5. Compulsory field
    1. [Error Message] Metric is required
  6. [Help Text] Select a metric to set up thresholds

Metric thresholds
– Overview

  1. Title: Metric Thresholds
    1. [Help Text] Configure up to 5 threshold ranges for your metric
    2. Compulsory field 
  2. By default, display the first threshold configuration as the Critical Range (Red). (Fig 3a, Fig 3b, Blue)
  3. Support '+ Add Threshold' button 
    1. Greyed out when the maximum number of conditions have been set up (Fig 4)
    2. Support a counter for the number of thresholds added

How to set up a metric threshold

  1. Choose a target metric
  2. Assign a severity level (Colours) to the condition of the target metric
  3. Set the data range for the selected severity level 
  4. Add more thresholds if required

**For this phase, conditions set up will apply to all data points. As of today, no requests have been made to specify a particular dimension with the value of the target metric.

Metric Threshold
– Severity Levels & Colours
  1. For each metric, user can add up to 5 metric thresholds
    1. In this phase, support only 1 configuration for each chart, even though a chart can have multiple metrics. 

  2. Each metric threshold has a distinct colour which signifies a different severity level associated with the target metric. (Fig 5)

    1. Only 1 severity colour can be used once per chart. 

    2. When 2 of the same severity colours are selected, highlight those input boxes with the same severity colour selected and display an error message beneath the dropdown box: 
      1. [Error Message] Same severity colour not allowed.
      2. User cannot click onto "Confirm", it will be inactive
        1. [Hover Text] Address errors before confirming
    3. By default, display the first configuration as Critical Range (Red) to show user that it is compulsory to at least have 1 threshold. (Fig 3a, Fig 3b, Blue)
      1. For every subsequent addition, the next severity level goes down the list to the next available one. Order of adding which severity next follows the table below. Example: 
        1. When user changes the first condition from #1 Critical (Red) to #5 OK (Green) then adds a new condition, the next condition will be #1 again. 
        2. When user changes the first condition from #1 Critical (Red) to #4 Low (Blue) then adds a new condition, the next condition will be #5. 

    4. Support colour gradient selection for chart types with conditional formatting.

    5. Rather than leveraging on the severity levels outlined in the BRD, SSAD will adopt more universal naming convention for our different pools of users. (Fig 5)

      Order
      Colour
      Severity Level
      #1RedCritical (Corresponds to Severity 4)
      #2OrangeHigh (Corresponds to Severity 3)
      #3YellowMedium (Corresponds to Severity 2)
      #4BlueLow (Corresponds to Severity 1)
      #5GreenOK (Corresponds to Severity 0)
Metric Thresholds
– Conditional Formatting
  1. Applicable to only Metric Card, Table & Pivot Table.
  2. Support up to 5 conditional formats (Fig 4)
  3. By default, display all conditional formatting set on the chart but user can hide or display it from the main page 
  4. For each severity, user can set a threshold range. 
    1. Display the title for each condition as '{Severity Level} Range' (Fig 4)
  5. Support deleting of condition
    1. If there is only 1 threshold range, no delete icon is supported (Fig 3a, Blue)
Subcomponent
Description
Severity Colour
  1. Support colour selection and option for either single colour or colour gradient
    1. Update the new shades for all 5 colours 
  2. Same as current conditional formatting but with 2 more colours available
Operator
  1. Support selection of operators: =, !=, >, >=, <. <=, between (inclusive) 
    1. Same as current conditional formatting
    2. [Placeholder Text] Select an operator
    3. Error Validation
      1. [Error Message] Required field
Threshold Value
  1. Support threshold value input
    1. [Placeholder Text] Input a number
    2. Error Validation
      1. For Between operator, return error message when the first value is bigger than the second value
        • [Error Message] Invalid range, first value must be smaller
      2. For empty values
        • [Error Message] Required field
  2. Display the selected number format of the metric from Data Tab
    1. When number format changes, the threshold value will remain the same. 
      1. E.g. When user sets up metric threshold with value 300min, then changes number format to h, the threshold value will not automatically convert to 5h. It will remain as 300 h
      2. Display a red dot on Monitoring tab to signal to user to check the threshold value since the number format has been updated.
        1. If the number format undergoes several changes but the user reverts it to the initially saved format for the threshold value, no red dot will be displayed.
      3. Highlight the threshold value that requires user attention. 

Multiple Condition Hierarchy (Overlapping Rules)

  1. Support display of information icon to inform user when a certain condition overlaps with another
    1. [Information] Overlap detected, priority follows: Critical > High > Medium > Low > OK (Fig 6)
    2. E.g. Overlap detected: Critical range takes precedence over High, Medium range
  2. When conditions overlaps one another, give priority to the higher severity level condition in the list. The other lower severity level ranges will still be applied if the high severity level range does not cover that range.  
    1. [AS-IS] Priority is given to the last condition in the list, however it is not clear to user as to which takes precedence. 
    2. The following scenarios are already implemented in SSAD: 
      1. If two ranges overlap, and the last condition's range (>30) doesn't fully cover the first condition's range (>1), the first condition will still cover values within its range (2-30) until the last condition's starting value (>30).
      2. When two ranges are exactly the same, the last condition (Yellow) will overwrite the earlier one (Red)
      3. When the last condition's range (>1) overlaps the first condition (>30) entirely, the last condition (Yellow) will overwrite the first condition (Red). 
Metric Thresholds
– Target Line

[AS-IS] User can display either a static target line or dynamic (average / median) target line, there were no conditions or colour selection required.

  1. Applicable to only Bar, Row, Stacked Bar, Line Chart. 
  2. Support up to 5 conditional target lines
  3. By default, display all target lines set but user can hide or display it from the main page 
  4. For each severity, user can set a condition for a target line. 
    1. Display the title for each condition as '{Severity Level} Range'
  5. Support deleting of condition 
    1. If there is only 1 threshold range, no delete icon is supported (Fig 3b, Blue)
  6. Remove dynamic target line
    1. With the low usage of dynamic target line, it is safe to sunset this function and instead integrate static target line and conditional formatting into 1 function. 
Subcomponent
Description
Severity Colour
  1. Support only single colour selection (no colour scale) 
  2. 5 new colour shades for the different severity levels
  3. By default, the target line is in grey. (Fig 7a) 
    1. When the value of the target metric hits the target line, target line will become coloured based on the severity colour
  4. When selected operator is >, >=, <. <=, between (inclusive):
    1. Cover the chart area with a lighter severity colour shade to show the range of values contained under that condition (Fig 7b)
Operator

Support selection of operators: =, >, >=, <. <=, between (inclusive) 

  1. != is not supported
  2. Comparing with the current static target line, it is equivalent to 'Greater or equal to (>=)' in Conditional Formatting because the line lights up when any value hits on or above the target line 
  3. E.g. Metric X Static, 400 →  Metric X >= 400 
Threshold Value
  1. Support threshold value input
  2. Same error validation as conditional formatting's
    1. For Between operator, return error message when the first value is bigger than the second value
      • [Error Message] Invalid range, first value must be smaller
    2. For empty values
      • [Error Message] Data range cannot be empty
  3. When there are same threshold values but the conditions do not overlap, display only the upper limit target line. (Fig 7c, 7d)
Confirm button

Inactive when:

  1. Target metric not selected
  2. Conditions are incomplete or when there no condition set up
  3. [Tooltip] Address incomplete fields before proceeding
Cancel button
  1. Support a cancel button to close the "Add Monitoring" panel
  2. User can also click outside of the panel, on the dashboard, to close the panel
  3. If there are configurations but user clicks onto cancel, display a pop-up
    1. [Pop-up] Are you sure you want to discard your configurations? 
    2. No – Stop the action
    3. Discard – Close the panel and discard all the settings

Fig 3a: Components of Monitoring Thresholds (Conditional Formatting)

Fig 3b: Components of Monitoring Thresholds (Target Line)

Fig 4: A maximum of 5 metric thresholds are allowed, each severity colour can only be used once

Fig 5: 5 severity colours available

Fig 6: Notice displayed when an overlap of threshold range is detected 

Fig 7a: Target line when threshold is not met (= 500) 

Fig 7b: Target line with shaded area to show range of values (between 100 and 390) 

Fig 7c: Target line for threshold range - >= 590, < 590

Fig 7d: Target line for threshold range - > 590, between 100 and 590

F012 – Existing Configurations 

For a seamless transition, replicate the configurations in the new function for those dashboards that have set up target line or conditional formatting.

  1. For those with Conditional Formatting configured: 
    1. Regardless of the colour type (single colour (Fig 8a)/colour gradient (Fig 8b)), 
      1. when colour = Red, replace with severity colour Critical Range
      2. when colour = Yellow, replace with severity colour Medium Range
      3. when colour = Green, replace with severity colour OK Range
    2. Carry over the same group metric as Target Metric and operator 
  2. For those with Target Line configured: 
    1. Carry over the same target field as Target Metric
    2. If Static, make operator as Greater than or equal to (>=) and retain the value as Threshold Value (Fig 1a)
    3. If Dynamic, no threshold range will be carried over since there is no equivalent. (Fig 1c)

Fig 8a: AS-IS Single Colour and its 3 different colours

Fig 8b: AS-IS Colour gradient and its 3 different colours

F02 – Alert Settings

Alert Setting consists of the following components: 

  • F021 – Check Frequency
  • F022 – Alert Frequency
  • F023 – Alert Notification (Seatalk)
  • F024 – Alert Notification (SSAD Alert Box Set-up)
  • F025 – Alert Notification (SSAD Alert Box Display)
  • F026 – Chart's Alert Label

ComponentDescriptionUI Reference
F02 – Alert Settings
  1. Alert setting is optional
    1. Support a toggle for user to on / off alerts (Fig 1a)
  2. There are several ways for user to set up an alert (Fig 1b): 
Component
Method 1 
Method 2
Method 3
Method 4
[Compulsory] Check Frequency (tick)(tick)(tick)(tick)
[Compulsory] Alert Frequency(tick)(tick)(tick)(tick)
[Compulsory] Alert Notification – Alert Box(tick)(tick)(tick)(tick)
[Optional] Alert Notification – Alert Box (Actions)(error)(tick)(error)(tick)
[Optional] Alert Notification – Seatalk(error)(error)(tick)

(tick)

ResultOnly alert box will be displayed on SSAD when triggered
  1. Only alert box will be displayed on SSAD when triggered
  2. Alert box contains suggested actions that include links to resolve the situation
  1. Alert box will be displayed on SSAD when triggered
  2. Alert notification will be sent to Seatalk group
  1. Alert box containing suggested actions will be displayed on SSAD when triggered
  2. Alert notification will be sent to Seatalk group

3. [IMPORTANT] Alert scope is based on dashboard owner's data permission level (All Company / My Company)

    1. Different users with varying data permissions (All / My Company) can access the same dashboard, with data objects filtered accordingly. This creates confusion regarding which data scope should be queried for triggering alerts. 
    2. Under the alert setting, support an info tag to inform users of this: 
      1. [Info] Alerts will be triggered based on dashboard owner's data permission level. {Dashboard Owner Name} has {All / My Company} data access. 

Fig 1a: Alert setting toggle OFF

Fig 1b: Alert setting fields when turned ON  

F021 – Range to be Alerted (Check Frequency) 

  1. Support a checklist of the threshold ranges configured in Metric Thresholds (Fig 3)
    1. Title: Range to be Alerted
    2. By default, there will be at least 1 threshold range because it is compulsory to have 1 range set
    3. User will select the threshold range(s) to send alerts for
      1. If there is only 1 threshold range configured, select the threshold range automatically
  2. Display "Check if metric is in alert range every X minutes/hours" 
    1. Support a numeric input box for user to define the evaluation period X of the target metric
      1. Minimum value is 5 mins
      2. Perform error handling upon immediate detection
        1. [Error Message] System performs checks at a minimum interval of 5 minutes.
    2. Support a dropdown for users to define the unit of time - minutes or hours
  3. Check Frequency pertains to how often the system examines or evaluates the metric thresholds to determine if an alert notification should be sent. It represents the rate at which the system actively reviews and assesses the specified criteria.
    1. Example: If monitoring occurs every 5 minutes, it means that at the start of every hour, the system will initiate counts at intervals of 5 minutes, such as 10:00 am, 10:05 am, 10:10 am, and so on.
  4. [Help Text] Select alert range(s) and define the frequency for system threshold monitoring. If the interval is every 5 min, it checks at the start of each hour at 10:00 am, 10:05 am, 10:10 am, and so forth.

Fig 2: Defining which threshold ranges to send alerts for and its check frequency

F022 – Alert Frequency

  1. Alert will only be sent if dashboard is in published state
    1. When dashboard is not in published state, prompt user to publish dashboard before alerts can be activated. 
  2. Support single select (Fig 3a):
    1. Send alerts when threshold is met (Always notifies if threshold is met at every check)
    2. Send alerts between {Start Time} - {End Time} (Notifies only during certain time period)
  3. [Help Text] Decide how frequent to be notify via Alert box or/and Seatalk. (Fig 3b)
    1. Always alert upon each interval check if thresholds are met.
    2. Limit alerts to specified time frames & alert upon each
interval check if thresholds are met.

Fig 3a: 2 different options for defining alert frequency

Fig 3b: Tooltip to describe the alert frequency options

F023 – Alert Notification  (Seatalk)

  1. Support single-select dropdown box for user to select which Seatalk group to send notification (Fig 4) 
    1. Optional field
  2. When selected threshold is met during interval check, send an alert notification to Seatalk group
  3. (TBC) List of Seatalk groups
  4. Seatalk Alert Content contains the following: 
    1. Number of criteria triggered - In a table, there may be multiple criteria that is triggered at a time
    2. Dashboard Name
    3. Chart Name
    4. For each alert, display:
      1. Criteria number
        1. If there is only 1 alert triggered, the number is not required
      2. An emoji corresponding to the severity level (Fig 5)
      3. {Metric Name} met {severity level} range of {threshold range}
      4. Recommended Action(s)
        1. Display the configured actions as a bulleted list
        2. If there is no actions configured, then hide this
    5. View dashboard now: 
      1. Dashboard link
    6. Sample (Fig 6): 

2 Alert(s) triggered

Dashboard: Seller Chat | L2 Team Cases v2
Chart: Agent AHT

#1 (warning) AHT met CRITICAL range of >= 5s

Recommended Action(s):

    • Delay agent break time

    • Increase max chat concurrency

#2 (red star) AHT met HIGH range of >= 10s

View dashboard now:

https://cs.shopee.co.id/portal/inhouse/web-ssar/dashboard/view/435

Fig 4: Dropdown to select seatalk group

Fig 5: Corresponding emojis to severity levels

Fig 6: Sample message sent on Seatalk



F024 – Alert Notification (SSAD Alert Box Set-up )

[AS-IS] In SSAD, user can display an action box in the chart itself, it contains an action description and up to 5 suggested actions in the form of hyperlinks. 

[TO-BE] Alert box shall give user information about what and how to note of any unusual behaviour from the target metric. Displayed on the top of the dashboard, it includes a description, the severity of the issue, and can include suggestions for how to resolve the situation.

  1. Alert box will appear when metric threshold is met
  2. Regardless of whether actions were mapped, alert box will still be displayed. 
  3. Remove action box from Customize Tab as it will be merged into Alert Setting (Fig 7)
SubcomponentDescription
Suggest Actions in Alert

Suggest Actions in Alert

  1. [Help Text] Suggest actions (up to 5 per range) for guidance on resolving issue. These actions will be displayed in the dashboard alert box (seen below) and in Seatalk notifications (if enabled) when metric thresholds are met.
  2. Optional field
  3. Support display of an image to show user how it looks like
Alert Message in Alert Box
  1. Display the alert boxes of the selected alert range (Fig 8a)   
    1. Default Message: {Metric Name}{Metric threshold}, refer to chart: {Chart Name}
      1. {Metric Name}: Selected metric name and it's aggregation method
        • E.g. CSAT (by Count) 
      2. {Metric threshold}: Operator, Threshold Value (with number format if any) 
        • Corresponding triggered conditions will be displayed when met
        • E.g. >= 10,000, between 100 and 200 
      3. {Chart Name}: Reflects the name of the chart
  2. Uneditable field 
    1. For this phase, just use default message and if needed, support configuration in next phase. 
Alert Actions
  1. Optional field
  2. Each alert range selected allows the user to configure up to 5 actions per range.
  3. Support '+ Action' button 
    1. Support multi-select dropdown box to choose up to 5 actions (Fig 8b)
      1. Refer to F021.1 action list below for available action options – Includes the existing 5 actions

    2. Hidden when all 5 actions have been configured.

    3. E.g. 4 actions were added, one more action can be added then it is hidden
  4. Support deleting of action configuration
Confirm buttonSame as above F011 Confirm and Cancel button 


Cancel button 

F021.1 Action List

There are a total of 28 new actions, change the url domain to the respective region's: 

#Action TypeAction NameLink
1

Maximize Current Capacity


Strictly manage non-online status

Agent Monitoring in Supervisor Console: 
https://cs.shopee.co.id/portal/inhouse/supervisor/monitor/agents



2

Increase max chat concurrency

3

Reallocate agent to another queue

4

Use templated answers

\ 
5

Delay agent break time

\
6

Request agents to be on standby for early log in / late log out

\

[Existing]
Check or change agent status

Supervisor Console: https://cs.shopee.XX/portal/inhouse/supervisor/monitor/agents

[Existing, to update] Add agents to queue

Queue Settings in IHS Configuration: https://cs.shopee.XX/portal/inhouse/routing-config/queues

[Existing] Change agent capacitySupervisor Console: https://cs.shopee.XX/portal/inhouse/supervisor/monitor/agents

[Existing] Change live agent controlChatbot Admin Portal: https://chatbot-admin.shopee.XX/chatbot/live-chat-traffic-control/config-overview

[Existing] Emergency switch to turn off live chatCS Admin Portal: https://admin.cs.shopee.io/portal/emergency-control
7Workforce Reallocation

Reallocate agent from other channel

Agent Monitoring in Supervisor Console: 
https://cs.shopee.co.id/portal/inhouse/supervisor/monitor/agents

8

Additional agent from L2

9

Request overtime working hours

\
10

Delay agent planned leaves / off schedule

\
11

Manage volume with low UX impact

Help Center announcement on high contact volume

Help Center Announcement: 
https://cs.shopee.xx/portal/inhouse/helpcenter/announcements  

12Chatbot announcement on high contact volume

Chatbot announcement:
https://chatbot-admin.shopee.com.xx/chatbot/announcements

13

Turn on high volume IVR flow

https://login.mypurecloud.jp/

14

Add audio message in IVR to inform on defined bug

https://login.mypurecloud.jp/
15

Force chatbot interaction by reduce show rate

https://chatbot-admin.shopee.com.xx/chatbot/live-chat-traffic-control/model-configuration

16

Update hot questions in chatbot for defined spike issues

https://chatbot-admin.shopee.com.xx/chatbot/display-orders/3x3

17

Manage volume with medium UX impact

Reduce inactive users (Zombie Chat Filter)

\

Please raise request to CS Reg Team

18

Traffic control based on queueing limit

https://cs.shopee.xx/portal/inhouse/routing-config/queues

19

Reduce reshow rate of live chat button after chatbot interaction

https://chatbot-admin.shopee.com.xx/chatbot/live-chat-traffic-control/model-configuration

20

Reduce duration of sending chaser (Auto Chaser)

\

Please raise request to CS Reg Team

21

Reduce duration to auto end chat after user inactive

\

Please raise request to CS Reg Team

22Manage volume with high UX impact

Targeted PN to users

\
23

Update delivery date in Shopee App

\
24

Remove phone number from Help Center

https://cs.shopee.xx/portal/inhouse/helpcenter/frontends

25

Extend operating hours during campaign period

https://cs.shopee.xx/portal/inhouse/case-config/workhour

26

Shorten operating hours

https://cs.shopee.xx/portal/inhouse/case-config/workhour

27

Traffic control based on handled contact limit

https://cs.shopee.xx/portal/inhouse/routing-config/queues

28

Emergency switch to turn off incoming contact

https://admin.cs.shopee.io/portal/emergency-control

Fig 7: Remove Action box 

Fig 8a: Suggest Actions in Alert, only the select ranges will require action configuration

Fig 8b: Adding actions


F025 – Alert Notification (SSAD Alert Box Display)

By default, alert box will be pinned at the top of the dashboard when conditions are set up. 

Subcomponent
Description
Alert Icon
  1. Display an alert icon with the severity colour (Fig 9a)
  2. Display the severity level – Critical, High, Medium, Low, OK 
Alert Message
  1. Display the alert message based on the triggered condition, follows F024
  2. Upon clicking the Chart name from the alert message, redirect user's view to the specific chart that triggered the condition 
Suggested Actions 
  1. Display a title: Suggested
  2. Display the suggested actions that user configured
  3. Support redirection via a new browser tab if configured action has a link 
    1. For actions with 'Please raise request to CS Reg Team', support a hover to display text.
Dismiss button
  1. Support a 'Dismiss' button to allow user to hide the alert box. (Fig 9b)
  2. Support two options: 
    1. Dismiss for today – Alert box will not appear on dashboard for the rest entire day, however notifications will still be sent to Seatalk if configured. 
    2. Dismiss this – Alert box will be hidden and displayed again during the next check
Navigation

User can set up an alert for every single chart on the dashboard, so there may exist multiple alert boxes. 

  1. Support a navigator to allow users to scroll 
  2. Latest and highest severity level trigger will be displayed as the first one of the stack 

Fig 9a: Alert box on SSAD

Fig 9b: Dismiss options

F026 – Chart's Alert Label

  1. Labels the chart that has an alert configured with an alert icon
  2. Support blinking effects when alert has been sent
  3. When alert is triggered, alert icon is coloured, otherwise, it will be grey.
  4. When alert box is dismissed, alert icon will return to grey colour
  5. Support tooltip over alert icon and display the list of configured check frequency:
    1. When alert is not triggered, display all the configurations: (Fig 10a) 
      1.  Checks every {check frequency} {min/h} if {Metric Name (Aggregation} is in: 
      2. {Severity Colour} {severity level} range {Threshold Range} 
    2. When alert is triggered display only the triggered alert info: (Fig 10b) 
      1. {Severity Colour} {Metric Name (Aggregation}  {Threshold Range} 
      2. Suggested Actions and its links
        1. Hidden if not configured

FIg 10a: All alert rules displayed when hover over grey icon beside chart title 

FIg 10b: Only the triggered alert rules is displayed when hover over coloured icon beside chart title 

F03 – View / Edit Configuration

View / Edit Configuration consists of the following components: 

  • F031 – View & Edit Configuration
  • F032 – Hide & Display 
  • F033 – Copying Dashboards and Charts
Component
Description 
UI Reference

F031 – View & Edit Configuration

  1. Support editing and deleting of metric thresholds and alert setting
  2. Display the target metric with it's aggregation method, severity colour and its threshold range
    1. Beside each threshold range, support a hide/display icon
  3. Display Alert Setting 
    1. Check frequency 
    2. Alert frequency
    3. Suggested Actions if any 
      1. When not added, support an + Add Suggestion button for user to easily map actions 
    4. Seatalk (if configured) 
  4. When user clicks edit, the "Confirm" button on the panel becomes a "Save" button instead. 

F032 – Hide & Display 

  1. For each threshold range configured, the target line or conditional formatting will be displayed on the chart by default. 
  2. Support an icon which allows user to show / hide any of them 
  3. There is no minimum to be shown or hidden
  4. This is independent of the alert setting, when hidden, alerts are still active

F033 – Copying Dashboards and Charts

Copying Charts with Monitoring and Alerts

Copied chart will carry over the configurations in monitoring and alerting

Copying Dashboards with Charts with Monitoring and Alerts

The charts in the copied dashboard will contain the same configurations in monitoring and alerting but data permission is now based on user who copied the dashboard. (New dashboard owner) 


F04 – Tutorial

Description 
UI Reference

F04.1 – Introducing Monitoring and Alerts 

Description: We have launched a new feature to monitor metric thresholds and send alerts to your Seatalk and dashboard! 


F04.2 – Configure Metrics to Monitor 

Description: Configure up to 5 threshold ranges based on your monitoring criteria. Conditional formatting and target lines setup have been migrated here to seamlessly align with alerting.


F04.3 – Send Alerts on Dashboard and Seatalk

Description: Get notified on both your Dashboard and preferred Seatalk group whenever metric thresholds are reached. Customise your alerts and suggest actions for your viewers to take.

With SSAD Alerts, you're always in control of when and how you stay informed!


F04.4 – Alert Box on Dashboard

Description: Once an alert is triggered, you'll be notified on the dashboard via an alert box which contains the chart name, metric thresholds and suggested actions if any. The same content will be available in the Seatalk alert if set up. 

When an alert is triggered, you'll get notified on the dashboard with an alert box displaying the metric thresholds, chart name and any suggested actions. Set up Seatalk alerts for the same updates!