Meiqia Intelligent Customer Service
Meiqia Intelligent Customer Service is an AI solution developed by Chengdu Meiqia Network Technology Co., Ltd. It is designed to meet the needs of corporate customer service, offering a range of intelligent services based on AI, such as intent recognition, emotion understanding, and context analysis.
Tags:AI training modelsWhat is Meqia Intelligent Customer Service?
Meiqia Intelligent Customer Service is an all-channel customer service platform designed for businesses, covering online consultations, AI robots, ticket handling, and call centers.
Customer service robots can answer common questions on their own, and they can also suggest answers when human agents are handling the inquiries, mark the conversations, and assist with summarizing them.
Main functions
- Common inquiries are handled automatically 24/7.
- Create an enterprise knowledge base and a pre-set knowledge base.
- Personalized responses are provided using the table knowledge base.
- Collect unidentified issues and keep learning.
- Recommend candidate replies to the human customer service agent.
- Automatically mark and summarize conversations.
- Configure different robots according to the channel.
- Access to web pages, apps, and social media platforms.
- Manage customer profiles and past conversations.
- Create and route service tickets.
- Configure marketing bots to collect leads.
- View data statistics and real-time dashboards.
- Integrated with call centers for handling telephone services.
AI customer service robot
The Meiqia customer service robot can handle inquiries independently during peak hours and at night; if it is unable to resolve a query, it will forward it to a human agent.
| Method of reception | Suitable scenarios | Configuration priorities |
|---|---|---|
| Robots first | There are many repeated questions. | Switch to manual mode and knowledge coverage |
| Human priority | High-value or complex consulting | Robot-assisted recommended answers |
| Night care | Consultations outside working hours | Leave messages, be on duty, and follow up the next day |
| Receive guests through different channels | Multiple brands or business lines | Channels are associated one-to-one with robots. |
| Human-robot collaboration | High requirements for service quality. | Manual verification with robotic fallback |
Corporate knowledge base
The enterprise knowledge base is used to store standard questions, similar queries, and their answers; it forms the basis for the quality of recognition and responses by robots.
- Collect real customer service conversations and common questions.
- Categorized by product, pre-sales, after-sales, and policies.
- Write a standard answer for each question.
- Add different phrasings that users might use.
- Set the time when the answer becomes valid and when it becomes invalid.
- Add the necessary links, images, or instructions to route to a human.
- Simulate various expressions in the testing environment.
- It will be continuously updated based on unidentified issues.
Table knowledge base
The table knowledge base is suitable for structured information such as prices, stores, courses, and order rules; it can provide more accurate answers based on user criteria.
| Data type | Example | Key maintenance points |
|---|---|---|
| Product information | Model, price, and inventory | Stay synchronized with business systems |
| Store information | Cities, addresses, and hours of operation | Deal with temporary changes in a timely manner. |
| Course services | Class types, schedules, and target audience | Avoid outdated enrollment information |
| After-sales rules | Deadline, conditions, and handling methods | Complies with the current contract policies |
| Event benefits | Time, requirements, and discounts | Set a clear expiration date. |
Intelligent learning
Problems that the robot cannot identify are added to the list for further learning; the administrator can add additional questions, create new ones, or ignore invalid content based on the context.
- Regularly open the list of unrecognized issues.
- View the frequency of the issue occurrence and the full context.
- Filter out garbled characters, advertisements, and meaningless input.
- Add synonymous expressions as new phrasings for existing questions.
- Create standard questions for genuine new requirements.
- Write answers that are executable and do not make excessive promises.
- Test it first before releasing it to the production environment.
- Continuously monitor the hit rate and the rate of cases being transferred to human agents.
Robot-assisted human customer service
Human agents can enable robot-assisted responses; when the knowledge base finds an answer to the visitor’s question, the robot presents possible responses for the agent to decide whether to send them or not.
- Reduce the time spent searching for information and copying scripts.
- Standardize the responses to common issues.
- A human determines whether the answer is appropriate for the current customer.
- Complex complaints cannot rely solely on automated responses.
- For refunds and contracts, it is necessary to check the real-time records.
- Incorrect recommendations should be fed back into the knowledge base maintenance process.
Multiple robots and multiple business lines
Enterprises can configure multiple robots, allowing each robot to use its own business card, knowledge base, and customer service channels.
| Business segmentation | Robot configuration | Advantages |
|---|---|---|
| Different brands | A separate robot for each brand | Avoid confusion between names and policies. |
| Pre-sales and after-sales support | Classified by service phase | The phrasing and handoffs are more accurate. |
| Different regions | Configure by language or region | Adjust to business hours and policies |
| Different product lines | Associate with the corresponding knowledge base | Reduce interference from irrelevant content |
| Websites and apps | Set up robots by channel | Provide different guidance based on the source. |
The help documentation states that each channel or sub-channel can be associated with only one robot; if a new robot is needed, it will have to be purchased separately.
Omni-channel online customer service
Meiqia enables the aggregation of inquiries from various channels into a single dashboard, thereby reducing the need for customer service staff to switch back and forth between different systems.
- Website plugins and standalone chat links.
- Android and iOS mobile apps.
- WeChat official account and WeChat customer service.
- WeChat Mini Program.
- WeChat Work accounts and platforms such as Douyin.
- Authorized channels such as REDnote professional accounts.
- The availability of different channels is subject to the workstation.
Customer data and conversation management
The customer service agent can view the customer’s tags, source, device, previous sessions, and tickets in the conversation window, as well as add the customer’s contact information.
- Use tags to distinguish intentions, questions, and risks.
- View the source keywords and visited pages.
- Save the necessary customer contact information.
- Maintain the service context through historical conversations.
- Convert unresolved issues into tickets.
- Add harassing visitors to the blacklist.
- Only collect the information necessary to complete the service.
Marketing robots
Marketing robots can ask questions to visitors through a visual interface, verify their responses, and collect information such as names, phone numbers, WeChat IDs, or email addresses.
- Identify the target audience and conversion goals.
- Design the welcome message and the first question.
- Set process branches based on user responses.
- Configure validation rules for name, phone number, etc.
- Write down the useful clues on the customer’s business card.
- Set options for exiting unfinished processes and transferring them to manual handling.
- Go through all branches using a test account.
- Check the data on leads and conversions after going live.
Ticketing system
The customer service staff can create a ticket from the issues discussed in the current conversation, and assign it to other employees or departments for further processing and tracking.
| Ticket processing stage | Main actions | Key points of inspection |
|---|---|---|
| Create | Record customers, issues, and evidence | The information is complete and contains no unnecessary private details. |
| allocation | Assign a department or responsible person | Responsibilities and priorities are clear. |
| Processing | Record progress and actions taken to resolve issues | Retain auditable processes |
| Reply | Sync the results with the customer | The caliber is accurate and feasible to implement. |
| Turn off | Confirm that the issue has been resolved. | Collect evaluations and consolidate knowledge. |
Call centers and voice robots
In addition to online customer service, Meiqia also offers telephone support and voice robot capabilities, which can handle incoming calls, outgoing calls, script management, and intent recognition.
- Manage agents, queues, and call records.
- Configure the script nodes for the voice robot.
- Identify the user’s intent and proceed to different branches.
- Format validation is used when collecting information.
- The key business settings are routed to manual processing.
- Comply with the rules regarding telemarketing and personal information.
- Set access permissions for recording and call data.
Tutorial for Setting Up Meqia Intelligent Customer Service
- Register for a corporate account and access the Meiqia workspace.
- Select the channels required for testing the trial version.
- Create a robot and set its name and avatar.
- Import common issues and corporate knowledge base.
- Configure welcome messages, recommended questions, and rules for transferring to a human agent.
- Associate the robot with web pages or SDK channels.
- Test knowledge accuracy using various phrasings.
- Test working hours, nighttime, and scenarios with no one online.
- Configure live agents, customer service teams, and assignment rules.
- Integration with ticketing and lead collection processes.
- After going live, monitoring detects no issues and provides data statistics.
- Regularly update knowledge and conduct random checks on service quality.
Tutorial for integrating web plugins
- Create a web access channel on the workstation.
- Obtain the web plugin code corresponding to the enterprise.
- Place the asynchronous script in the website template.
- Set the style of the chat window by brand.
- Configure welcome messages and automatic messages.
- Pass on the necessary user-defined information.
- Test the computer and mobile phone pages separately.
- Check the privacy notice and consent process.
- After deployment, observe the loading performance and conversation data.
Mobile app SDK
The authorities provide SDKs for Android and iOS, which allow chat interfaces to be integrated within apps and enable linking of app users with Meiqia customers.
| Platform | Connected content | Precautions |
|---|---|---|
| Android | Initialize the Appkey and start the conversation. | Configure dependencies and image loading |
| iOS | Integrated SDK and chat interface | Configure permissions and network policies |
| APICloud | Use the official adaptation module. | Check compatibility according to the platform version. |
The current public versions available in the official Android repository can be used as a reference for integration, but before going live it is necessary to rely on the latest SDKs provided in the help center and the dashboard.
Price and version comparison
The official website of Meiqia does not currently display the current prices of its various paid versions in a stable manner; therefore, the prices listed in past blogs cannot be used as a basis for making purchases today.
| Plan | Current public price | Robot messages | Main positioning |
|---|---|---|---|
| Trial version | 0 yuan | 1000 items are given as a gift. | Basic experience and channel testing |
| Professional Edition | Contact us for a business inquiry | 1000 items are given as a gift. | Online customer service for small and medium-sized teams |
| Enterprise Edition | Contact us for a business inquiry | Contains 100,000 entries | Robots, multiple channels, and advanced capabilities |
| Platform version | Request a quote after signing the agreement | In accordance with the contract | To enable bulk connection for enterprises under the platform. |
| Privatization or customization | Contact Business | According to the plan | Large organizations and complex integrations |
When a robot answers a question that matches a predefined pattern, it consumes message slots; no slots are used if the question does not match or if no answer is given. If the available quota is insufficient, it is necessary to contact the business team to purchase more slots.
How are robot messages calculated?
- All robots share the total number of remaining messages in the enterprise.
- Matching the problem and providing an answer is considered an act of consumption on the part of the accountant.
- No consumption occurs when no matching issue is found.
- The trial version and the professional version come with 1,000 messages included.
- There are 100,000 entries in the enterprise version’s help documentation.
- After it is depleted, you need to contact business support for top-up.
- Different contract versions may have varying limits and rules.
How to compare before making a purchase
| Dimension | Statistics are needed. | Procurement issues |
|---|---|---|
| Number of inquiries | Average and peak daily conversations | Are the robot messages and agents sufficient? |
| Channels | Websites, apps, and social media platforms | Does the target channel support it? |
| Agent | Concurrency and number of accounts | Billing by agent or by account |
| Robot | Number of business lines and knowledge bases | How is charging handled for new robots? |
| Ticket | Departmental structure and complexity of processes | Is a specialized ticketing module needed? |
| Integration | CRM, order, and identity systems | API, platform version, and implementation costs |
Data statistics and operations
Robot analytics can help businesses understand their volume of inquiries, success rates, and the performance of various channels, while data dashboards are used for real-time monitoring of service status.
- Observe the robot’s ability to solve problems independently on a proportional basis.
- Statistics on unidentified issues and frequent themes.
- Analyze the handover rate and wait time.
- Compare the volume of inquiries from different channels.
- Randomly inspect cases with low satisfaction and repeated inquiries.
- Use data to determine the priority for updating the knowledge base.
- Business metrics should take into account actual transactions and the resolution rate.
API and platform version
The platform version is intended for companies that need to enable the Meiqia service for a large number of their clients, and it uses Public Key and Secret Key to sign API requests.
- For the platform version, it is necessary to first sign a usage agreement with the business party.
- The key must be stored on the server side.
- A timestamp and HMAC signature are required for the request.
- It is possible to create enterprise accounts and customer service accounts under the platform.
- It is necessary to save the mapping between enterprises and customer service identifiers.
- Different enterprise plugins can be pre-installed in the web page template.
- The production environment should include retry mechanisms, logging, and rate limiting.
Official GitHub and open-source scope
The official GitHub repository of Meiqia provides various mobile SDKs and integration examples, but the Meiqia SaaS platform as well as the AI robots themselves are not open-source projects.
| Project | Public information | Uses |
|---|---|---|
| Android SDK | Public on official GitHub | Integrate customer service within the Android app |
| iOS SDK | Public on official GitHub | Integrate customer service within the iOS app |
| APICloud module | Public on official GitHub | Suitable for APICloud applications |
| Meiqia SaaS Platform | Not open source | Online customer service and enterprise management backend |
| Customer service robot model | Unpublished weights | Managed AI services |
Privacy and security
- Only the customer information necessary to provide the service is collected.
- A clear privacy notice is displayed before clue collection.
- Set minimum permissions for agents and administrators.
- The accounts of employees who have left the company should be deactivated promptly.
- API keys and Appkeys are managed separately according to their purpose.
- Sensitive fields are masked in logs and exports.
- Regularly review third-party channel authorizations.
- Set retention periods for phone call recordings and conversation logs.
Which companies are suitable?
| Companies | Typical requirements | Recommendation capability |
|---|---|---|
| E-commerce and retail | Pre-sales, order, and after-sales inquiries | Robots, online customer service, and ticketing |
| Education and training | Course consultation and lead collection | Marketing robots and spreadsheet knowledge base |
| Software and the Internet | Product support and APP integration | SDK, knowledge base, and tickets |
| Chain and local services | Store locations, booking, and service policies | Table knowledge base and multiple robots |
| Platform-based enterprises | Provide customer service for its merchants. | Platform API |
Product advantages
- Robots and human agents work together.
- Covers web pages, apps, and various external channels.
- The knowledge base, intelligent learning, and data statistics are all complete.
- It includes ticketing, marketing, and call center modules.
- It provides web plugins, SDKs, and platform-based APIs.
- The official GitHub continues to provide the mobile SDK.
Usage restrictions
- The current public page does not display the current price consistently.
- The prices for historical blogs are outdated.
- There are limits on the amount of messages that robots can send.
- Additional robots and excess messages may incur separate fees.
- The quality of the knowledge has a direct impact on the responses provided by robots.
- Complex complaints and high-risk transactions still require manual handling.
- The open-sourcing of the official SDK does not mean that the entire platform is open-source.
Frequently Asked Questions
What are the main functions of Meqia’s intelligent customer service?
It offers AI customer service robots, an enterprise knowledge base, intelligent learning tools, assistance from human agents, multi-channel access, ticketing systems, marketing functions, and a call center.
Can Meqia be used for free?
Newly registered companies are set to the trial version by default, allowing them to test the basic functions first; the full set of features, time limits, and seat limitations are specified in real time on the dashboard.
How much does the Meqia customer service robot cost?
The official website does not currently display the current price consistently; the trial version and the professional version come with 1,000 messages, while the enterprise version includes 100,000 messages. For any additional messages, contact the sales team.
How are the fees for Meqia robots deducted?
When a visitor’s question is matched by a robot and an answer is provided, one unit is consumed; no unit is consumed if the question is not matched or no answer is given.
Which connection methods does Meiqia support?
It can be accessed via web pages, Android, iOS, WeChat official accounts, WeChat customer service, mini-programs, as well as certain content and short-video platforms.
Does Meiqia provide APIs and SDKs?
It provides web JavaScript plugins, Android and iOS SDKs, as well as a platform API for use after signing a contract.
Is Meqia Intelligent Customer Service an open-source project?
The Meiqia SaaS platform and its robots are not open-source, but the official GitHub page provides SDKs and examples for clients such as Android, iOS, and APICloud.
Guigong Network Security Registration No. 45132202000164