What is Xiaojinglong
Xiaojinglong is an AI intelligent learning companion (Agent) for students launched by Xueersi, built on theOpenClawarchitecture. Through a "brain mapping," the tool continuously understands students' knowledge gaps and thinking habits, precipitates their learning trajectories into "private property" (the more it is used, the better it understands the student), and provides personalized tutoring and emotional companionship in the role of a "learning companion." Xiaojinglong can adjust its interaction style for different ages and personalities, proactively manage learning progress, and identify emotional states. Its positioning is a dedicated AI Companion that accompanies students' growth 24/7.
Main features of Xiaojinglong
- Learning diagnosis: Continuously analyzes learning trajectories to build a dynamic knowledge graph, accurately locates knowledge gaps and ability boundaries, and predicts the next optimal learning action.
- AI problem explanation: Supports taking photos and uploading homework or test papers, and provides step-by-step detailed analysis and in-depth thinking guidance. It not only gives answers but focuses more on explaining the problem-solving approach.
- Memory precipitation: Permanently saves all learning data, including incorrect questions, preferences, and conversation records, forming a personal exclusive cognitive asset and achieving a personalized experience that understands the student better the more it is used.
- Emotional companionship: Dynamically adjusts the interaction style according to the student's age, personality, and learning status, providing emotional support and learning motivation while offering academic tutoring.
- Skill configuration: Decouples educational capabilities into Skill modules that can be flexibly invoked, supports on-demand assembly of different functions such as diagnosis, problem explanation, and calculation, and allows setting scheduled task reminders.
- Convenient interaction: Provides an Android App that is ready to use upon download, supports multimodal interaction methods such as voice, photos, and text, and enables intelligent learning without complex configuration.
How to use Xiaojinglong
- Download and install: Visit the Xiaojinglong official websitehttps://mate.tal.com/下载App。
- Quick login: After installation, log in with one click by scanning a WeChat QR code, and directly use all functions without a cumbersome registration process.
- Ask questions and interact: Ask the AI learning questions in three ways: by taking photos and uploading homework or test papers, by pressing and holding the voice button to speak, or by manually typing text.
- Get explanations: The system automatically recognizes the question content and generates step-by-step detailed explanations. You can select "deep thinking" mode to view the complete derivation process.
- Learning management: View your personal knowledge graph and mastery analysis in the "brain mapping" feature. Continuous use allows the system to precipitate exclusive data and become more accurate the more it is used.
- Task configuration: Flexibly configure Skill modules such as diagnosis and problem explanation according to learning needs and set scheduled reminders to establish personalized study plans and progress management.
Key information and usage requirements of Xiaojinglong
-
Product positioning:An AI intelligent learning companion (Agent) for students launched by Xueersi (TAL), positioned as a "human-level intelligent learning companion," belonging to the same "Lobster" product matrix as the teacher-side "Jiuzhang Lobster."
-
Technical architecture: Built on the OpenClaw framework, with deep education-native transformation, setting education-specific guardrails to reduce hallucinations and decoupling capabilities into invocable Skills.
-
Core features: Brain Mapping” (dynamically understanding knowledge gaps), "Private Property" (accumulating exclusive learning data), "Learning Companion" (dual role of tutoring + companionship).
-
Usage threshold: Minimal deployment, "instant understanding," no need to configure an environment or write instructions to get started.
Xiaojinglong's Core Advantages
- Education-native Agent: Deeply transformed based on OpenClaw, with education-specific guardrails set up, focusing on learning scenarios to reduce hallucinations in general-purpose large models; compared with general AI tools, it is more stable, controllable, and precise in output.
- Deep learning-status understanding (Brain Mapping): Continuously builds a dynamic knowledge graph, sees through the knowledge gaps and cognitive biases behind problems, and predicts the next optimal learning action to achieve true personalization.
- Long-term memory accumulation (Private Property): Breaks through the limitation of traditional AI that "memory is lost as soon as the conversation ends," permanently preserving wrong-answer trajectories, thinking preferences, and conversation records to form exclusive cognitive assets; the longer it is used, the more accurately it adapts to the student's level.
- Integrated tutoring and companionship (Learning Companion): Dynamically adjusts interaction methods according to the student's age, personality, and emotional state, providing emotional support and learning motivation beyond academic tutoring, taking on a 7×24-hour companionship role.
Comparison of Xiaojinglong with Similar Competitors
| Dimension | Xiaojinglong(TAL Education) | Xiaoyuan AI(Yuanfudao) | Doubao AiXue(ByteDance) |
|---|---|---|---|
| Product positioning | Exclusive AI learning companion (Agent) | Problem-solving tool + AI Q&A | General AI assistant + education scenarios |
| Core capabilities | Brain mapping + companionship + tutoring | Question bank matching + step-by-step analysis + video | Large model dialogue + knowledge Q&A |
| Memory mechanism | Long-term private property(accumulates personal wrong answers, preferences, and trajectories, continuously evolving across sessions) | Short-term dialogue/history record viewing | Short-term context/no continuous accumulation |
| Personalization depth | Dynamic knowledge graph(predicts the optimal learning path, proactively pushes) | Similar-question recommendations based on the question bank | Relies on prompts for temporary adjustments |
| Emotional companionship | Proactive emotion recognition(adjusts interaction methods according to state, psychological support) | None/limited to problem-solving encouragement | General dialogue, no education-exclusive companionship |
| Technical architecture | OpenClaw education-native(capabilities decoupled into Skills, education guardrails prevent hallucinations) | Self-developed large model + question bank | Lark large model + general capabilities |
Xiaojinglong's Application Scenarios
- Homework tutoring scenario: When students encounter difficult problems, they take photos and upload homework or test papers. AI automatically recognizes them and explains the problem-solving approach and derivation process step by step in detail, replacing the traditional question-search model that only provides answers.
- Pre-exam diagnosis scenario: Through the "Brain Mapping" feature, view your personal knowledge graph to intuitively understand your mastery and weak points in each chapter. The system predicts the optimal review path based on historical data, enabling precise identification and filling of gaps.
- Wrong-question management scenario: Automatically categorizes and accumulates past wrong questions and error-prone knowledge points, regularly pushes targeted consolidation exercises, avoids repeating mistakes, and forms a personal专属 "wrong-question asset library."
- Study planning scenario: Based on the results of learning diagnosis, intelligently generates daily study tasks and review plans, sets timed reminders to manage study progress, and proactively pushes reminders when procrastinating or distracted.
- Emotional companionship scenario: When experiencing learning anxiety, frustration, or wanting to give up, AI recognizes the emotional state and provides appropriate encouragement and psychological guidance, adjusting the communication style (direct or gentle) according to personality traits to stabilize the learning rhythm.