Tag: learning

I-Bricks unveils next-generation AI tutor to enable personalized, self-directed learning (AI-generated image) | Image provided by I-Bricks

AI technology company I-Bricks is launching an advanced AI tutor system designed to deeply analyze learners’ cognitive states and encourage proactive conceptual exploration, aiming to establish a truly personalized and self-directed learning environment.

While existing AI learning services focus primarily on content recommendation or repetition-based practice, I-Bricks differentiates its technology by integrating a proprietary large language model (LLM) with deep knowledge tracing (DKT) to promote learner autonomy.

The system analyzes students’ problem-solving histories and conceptual understanding in real time, helping them identify recurring mistakes or areas of confusion. This enables students to check their own understanding and adjust their learning approach accordingly, fostering metacognitive learning activities.

To enhance metacognition, I-Bricks incorporates two key AI-based technologies into its learning system.

The first is a deep learning-based knowledge tracing model that monitors students’ problem-solving data and tracks changes in comprehension in real time. This allows the system to pinpoint areas where students struggle and prioritize concepts that need reinforcement.

The second is an error analysis model based on learning logs, which examines mistake patterns and cognitive processes by analyzing data such as problem-solving sequences, changes in answer choices, and time distribution. This model identifies areas of confusion and repeated errors, allowing targeted intervention.

Together, these technologies form the foundation of the I-Bricks AI tutor, which is designed to help students explore and understand what they don’t know and why they make mistakes. This empowers learners to actively adjust their learning paths.

The tutor also incorporates a personalization algorithm that provides motivational feedback based on real-time performance data, helping students stay engaged. Additionally, I-Bricks uses a large multimodal model (LMM) to summarize and present learning materials—text, videos, and slides—improving comprehension and focus.

At the end of each learning session, a reflection algorithm prompts students to look back on their learning with questions like “How could you solve this next time?” This encourages self-assessment and deeper understanding. A personalized capability graph visually presents changes in performance and highlights strengths and weaknesses, helping learners track their growth and gain motivation.

CEO Jonghyun Chae stated, “Our ultimate goal is to help every learner realize their full potential through AI tutors, equipping them with the self-directed learning skills needed in today’s world.”

Currently available for elementary, middle, and high school students, I-Bricks’ AI tutor will continue to evolve through ongoing technological innovation and user experience enhancements.

The company has also built a strong foundation in the edtech sector. Since 2017, I-Bricks has successfully developed and operated EBS’s AI-based chatbot tutoring service “Danchoo,” now in its fourth iteration. This project laid the groundwork for the current AI tutor, featuring a knowledge tracing model that predicts learning outcomes and offers personalized recommendations across all school subjects.

Founded in 2016, I-Bricks is a Korean language processing company specializing in hybrid chatbot development, real-time content recommendation based on user behavior, and unstructured data analysis using text analytics. The company is also actively developing compact large language models for generative AI applications.