If you want to know how to install LingChat and how to use the basic functions, please visit：LingChat Product Manual V1.1 20250331.
Through this article you will learn:
List of LingChat Bot Capabilities and Case (Part 1 and 2)
How to Cooperate with Human Agents to Improve Reception Efficiency and Manage Response Risks (Part 3 and 4)
How to Optimize Bot Operation Through Case and Data Analysis (Part 5 and 6)
How to Configure and Enable LingChat Bot (Part 7)
1.What LingChat Bot Can Do #
1.1 Product Recommendation #
Case: Buyers want to get a ribbon suitable for little girls recommended , the bot gives advice and supports clicking the card to jump to the product details page.
Q：Recommend a ribbon suitable for a little girl
Buyers consulting page (ask a question ➡ get the recommended content ➡ view recommended products and terms ➡ jump to the product details ➡ continue the chat)
1.2 General Knowledge Response #
Case: Consulting about return and exchange information, providing operation guidelines and introducing the information required from the buyer.
1.3 Chatting Reception + Question Clarification #
Case: Express service attitude professionally, guide customers to ask questions, and also block some invalid questions from customers, so that human can focus on solving complex problems.
1.4 Customer Service Risk Control: Customers’ Anger or Complaints + Inability to Answer Questions, Automatically Transfer to Human Customer Service. #
Case: When customers express complaints and impatience, automatically transferred to human customer service (the transferring words can be configured by the enterprise). User-side page (express complaints and impatience 
human reception)
1.5 Multi-language Response #
Case: The knowledge base is maintained in Thai. When the customer asks a question in English, the bot can answer it directly.
2.List of LingChat Bot Capabilities #
LingChat can prioritize responding by bots when aggregating reception, and the system will automatically transfer to human when the bots can’t respond, so as to solve complex problems by human and improve the efficiency of reception.
2.1 Channels and Sites Supported by Bot Response Capabilities #
| Platform | Thailand | Malaysia | Singapore | China Seller | China Taiwan |
|---|---|---|---|---|---|
| lazada | ✅ | ✅ | ✅ | / | |
| shopee | ✅ | ✅ | ✅ | ✅ | ✅ |
| tiktok | ✅ | ✅ | / |
2.2 Platforms and Specific Capabilities Supported by Bot Response #
| Capability Points | Capability Description | Lazada | Shopee | TikTok |
|---|---|---|---|---|
| Basic data-customer service messages | Support for bots response through official customer service messaging outlets | ✅ | ✅ | ✅ |
| Basic data – product data | The system automatically obtains product data after store authorization. | ✅ | ✅ | ✅ |
| User inputs recognized by the bot | Texts | ✅ | ✅ | ✅ |
| QA Knowledge Response | Respond based on QA knowledge uploaded by the enterprise and associated with the bot. | ✅ | ✅ | ✅ |
| Product recommendation | Based on the purchase intent described by the user, the bot will recommend products from the store that have been acquired in advance. | ✅ | ✅ | ✅ |
| Thank you for ordering | After the user expresses that the order has been placed, the bot expresses its gratitude. | ✅ | ✅ | ✅ |
| Polite response | Respond to the user’s thank you when the user expresses his/her gratitude. | ✅ | ✅ | ✅ |
| Multi-language response | Recognize the language of the user and respond in the language of the user. | ✅ | ✅ | ✅ |
3.Understand How to Transferring from Bot to Human #
In the case of abnormal customer moods, bots being unable to respond, or poor quality robotic responses, the human agent can intervene in a timely manner to ensure service quality and satisfaction.
| Capability Points | Capability Description | lazada | shopee | tiktok |
|---|---|---|---|---|
| [Transfer Method] Bot automatically transfer | When the bot cannot respond, it stops reception and sends words of transferring to human | ✅ | ✅ | ✅ |
| [Transfer Method] The agent takes over on its own initiative | The agent clicks the takeover button to take over the bot chat message and a human continues to receive. | ✅ | ✅ | ✅ |
| [Transfer Rules] Emotion recognition transferring | Automatically stops the bot reception when the user’s emotion is recognized as anger or complaint. | ✅ | ✅ | ✅ |
| [Transfer Rules] Unable to answer transferring | When the bot is unable to answer, it stops receiving and does not reply to the user’s message. | ✅ | ✅ | ✅ |
| [LingChat Message List Logo] Bot Response Logo | If it is a bot service state, the bot logo will be displayed, and the bot logo will disappear after transferring to a human | ✅ | ✅ | ✅ |
| [LingChat Message List Logo] Abnormal customer mood logo | If the customer’s emotion is abnormal and cause to transfer to human, the abnormal customer emotion logo will be displayed on the message list. | ✅ | ✅ | ✅ |
3.1 Bot Automatic Transferring Rules #
By default, the bot will automatically transfer to human when it can’t respond and the user’s emotion is abnormal (angry + complaining).
3.2 LingChat Message Logo + Agent Active Takeover Button #
4. How to Check the Bot Response Result and Control the Risk inTime through LingChat? #
If you want to know how to install LingChat and how to use the basic functions, please visit: LingChat Product Manual V1.1 20250331.
4.1 LingChat Page When the Bot is Working #
Human customer service can check whether the bot is working through the message list on the left
Human customer service can check whether the reply effect ofthe bot is OK or not.
4.2 Human Takes over the Bot Chat when Necessary #
Human customer service can take over the bot to reply manually by clicking the take over button at the bottom right corner of the page.
Human customer service can also withdraw the message replied by the bot and reply manually.
4.2 Prioritize Customer Chats that are Transferred to Human Due to Abnormal.Customer Mood #
(Bot response logo System automatically transfers to human chat Message list displays the angry logo)
View unreplied messages through the pending replies list
查看客户情绪logo，优先进行关注和处理
5. How to Optimize the Response Effect of the Bot #
The following operations need to be performed through the bot management backend, you need to login to the web version first:https://shopmate.3win.ai/login (account name and password are the same as LingChat)
5.1 Refining Generalized Bot QA Knowledge Response Capabilities throughCase Analysis #
Step 1:Check Bot Response through Chat History Viewing #
Link：https://shopmate.3win.ai/dataDashboard
Step 2:Conduct performance feedback by Like and Dislike #
Step 3: Refine QA knowledge #
Link：https://shopmate.3win.ai/source/knowlage/qa
Step 4: Edit and Update QA Knowledge #
5.2 View Overall Response Data through Bot Dashboard Analysis #
View bot response data
Link：https://shopmate.3win.ai/dataDashboard
5.3 View Human + Bot Response Data via Lingchat Dashboard Analysis #
The following operations need to be performed after log in the web version LingChat management backend: https://lingchat.3win.ai/login (account name and password are the same as LingChat client side’s account and password).
Please make sure your account and password has administrator privileges.
Part1：Today’s Monitoring #
Provide the basic data of LingChat reception on the same day to help administrators quickly grasp the receptionsituation on the same day (support manual refresh to view the real-time data + support filtering by platform/storedimensions to view).
Part2：Data Analysis-Conversation Trends #
View the trend of customers and messages served over time by day (support filtering by date/platform/storedimensions to view)
Part3：Data Analysis-Visitor Channel Distribution #
View the distribution of visitor inquiries in different stores, and visualize the proportion of customer inquiries instores (support filtering by date/platform’store dimensions to view)
Part4：Data Analysis-Top 10 Quality Tags #
View the Top 10 positive and negative tags summarized by AI to visualize the quality of customer service (support filtering by date/platform/store dimensions to view).
Part5：Data Analysis-Detailed Data #
Support exporting all detailed data by date (days) and store dimension (up to 30 days).
6.How to Configure and Enable LingChat Bot #
Step 1: Log in to the web version of LingChat management backend #
You need to log in to the web version to authorize the store interface and configure the bot
Link：https://lingchat.3win.ai/login
Step 2:Go to Settings -Channel Docking Page #
Step 3:Click Docking Channel #
Complete the authorization of the corresponding bot message interface of the store
Step 4: Jump to the Authorization Interface of lazada (take Lazada as anExample) #
Enter the site, account and password of the platform store and complete the authorization
Successful authorization will automatically jump to LingChat backend
Step 5: Go Back to LingChat Page to See the Successfully Bound Stores in the Docking List #
Successfully authorized stores can be found in the list
Click the right [settings] button to go to the bot management backend to configure the bot
Step 6: Click Settings to Automatically Jump to the Bot’s Management Backend Page #
In this page, you can view the product knowledge obtained from the store and the QA knowledge associated with the bot.
On this page, you can also test the bot to verify its effectiveness, and after passing the test, you can start the production environment.
Step 7: Upload the QA Knowledge through the Knowledge Base and Associate itwith the Bot, and the Bot will Respond According to the QA Knowledge #
Download the template and import the QA knowledge via Excel
Associate the imported QA knowledge
Step 8: Enable and Disable the Bot in LingChat Management Backend #
Step 9: Log in to LingChat to View the Effect of Bot Response #
If you encounter any problems in the process of using Ling Chat, you can contact us through the following ways:
Questions consultation email：product-rd@3win.ai
Questions consultation WhatsApp：+66 955018417
