Solo AI product builder.
Products people
come back to.

A Mandarin course for Vietnamese speakers that runs from HSK 1 to HSK 7–9 on the HSK 3.0 standard. It uses Sino-Vietnamese (Hán Việt) readings as a memory bridge, and Tiểu Long, a baby dragon mascot, guides the way.

• Smart flashcards: spaced repetition with tone-coloured pinyin, Hán Việt readings, Vietnamese meanings and example sentences, plus a searchable dictionary by hanzi, toneless pinyin, Hán Việt or Vietnamese.
• Full curriculum: stroke-order writing practice, the 214 radicals, grammar notes flagging mistakes Vietnamese learners often make, and HSK mock tests for each level with scoring and explanations.
• Speak it out loud: scenario-based AI conversations with corrections in Vietnamese, and shadowing practice that records the learner, transcribes their speech and compares it with the model sentence.
• Habit loop: imperial-exam titles from Đồng sinh to Trạng nguyên, streaks, review games, a mistake log and a weekly group leaderboard.

IELTS prep for Vietnamese learners, from band 4.5 to 9, coached by Jack, an orange cat the app also calls Mít. It is built around honest progress, so it never gives a flattering band score.

• Objective placement: an adaptive vocabulary test with grammar, timed reading and optional listening, weighted and rounded down. Band estimates are split by skill, and the app says plainly when it doesn't have enough data yet.
• Plan that fits the day: a daily route built from the target band and minutes per day. Listening uses a distinct natural voice for each speaker plus dictation, and the full 40-question Reading mocks convert to a band.
• AI feedback: Writing and Speaking scored against the four IELTS criteria, with Speaking recorded on phone or desktop. An AI room lets learners debate, practise paraphrasing and turn real English text into flashcards.
• Habit loop: titles from Freshman to Professor, streaks, daily tips, pronunciation drills, games and a mistake log.

GMAT Focus Edition prep covering Quant, Verbal and Data Insights, for complete beginners and 705+ targets alike. Max the fox is the study buddy and AI tutor.

• Exam-real practice: all 8 question types, including sortable Table Analysis, SVG-drawn charts, multi-tab Multi-Source Reasoning and an on-screen calculator for Data Insights. There are 5 difficulty levels, timed modes, and a worked solution with trap notes for every question.
• Adaptive mocks: a full 64-question exam with a choice of section order, a 10-minute break and Review & Edit for up to 3 answers. There is also a 15-question diagnostic.
• Transparent scoring: section (60–90) and total (205–805) estimates from a 3-parameter IRT model, clearly labelled as an estimate rather than GMAC's algorithm.
• Learning loop: an 81-step "Start here" path and an error log that tags why each question was missed. Max gives up to 3 step-by-step hints without revealing the answer. Titles run from Intern to CEO.

Learn AI and robotics from zero, deep enough to understand the field and build a career in it: AI engineering, forward-deployed engineering, robotics or AI product. Lessons are explained in Vietnamese with English terminology, and Bit the robot keeps learners company.

• Six-track skill tree: Programming, Math for AI, Machine Learning, LLM & GenAI, Robotics and Career. Each lesson comes with key terms, a 4-question quiz, an auto-graded coding exercise and a linked lab.
• Code in the browser: exercises run right in the page, auto-stop after 12 seconds to prevent runaway loops, and explain errors in Vietnamese. There is a free-form playground too.
• 17 interactive labs: tokenizers, temperature and top-p, embeddings, attention, a mini RAG, gradient descent, backprop, robot-arm kinematics, drone PID, A*, Q-learning, Kalman filters and lidar.
• Neuralis Arcade: the first version was too reading-heavy, so learning games, achievement badges and weekly quests were added as a retention lever: a reason to come back, not just another page to finish.
• Career map: 14 AI and robotics career paths, each with a day in the life, skills, tools, a starter path, portfolio ideas and interview questions. Bit, the AI assistant, can review the learner's code and plan a route to the chosen path.