HintWise

Inspiration

HintWise was inspired by a common learning moment: a student gets a question wrong, looks up the answer, and still does not understand why their reasoning failed.

Many study tools are optimized for quick answers. HintWise is designed for the step before the answer: helping students slow down, explain their thinking, receive a useful hint, and build real understanding from mistakes.

What it does

HintWise is a bilingual Socratic learning coach that helps students turn confusing questions into guided learning loops.

Students can enter a question, choose a subject and grade level, and start a tutoring session. HintWise analyzes the question, identifies the key concept, estimates the difficulty, and highlights a likely misconception.

The tutor then asks what the student tried first. Instead of revealing the final answer immediately, HintWise gives progressive hints. After the student makes an attempt, it unlocks a step-by-step explanation, generates three related practice questions, and saves the original question into a review notebook.

Key features include:

• Question analysis
• Socratic tutor chat
• Progressive hints
• Explanation unlocked after a student attempt
• Three follow-up practice questions
• Review notebook with concept tags
• English, Chinese, and bilingual learning modes

How we built it

HintWise was built with MeDo as a focused educational web app. The product flow was designed around one complete learning cycle:

Question → Analysis → Tutor Hint → Student Attempt → Explanation → Practice → Review

The interface is organized into four main sections: Workspace, Tutor, Practice, and Review Notebook. This structure keeps the experience simple for students while still supporting a complete learning journey.

AI is used across the core experience: analyzing the question, guiding the tutoring conversation, adapting explanations to grade level, generating related practice questions, and creating review cards.

Challenges

The main challenge was designing HintWise to be helpful without becoming an instant answer machine. The goal was to support learning, not shortcut it.

That meant carefully shaping the tutor behavior: asking for the student's first attempt, giving one hint at a time, and unlocking the explanation only after the student has engaged with the problem.

Another challenge was making the app useful across different subjects and languages. HintWise supports math, science, language arts, history, and bilingual learning, so the interface and tutor responses needed to stay clear, flexible, and student-friendly.

What we learned

Building HintWise showed that educational AI is not only about producing correct answers. The learning experience depends on timing, guidance, and reflection.

A small hint at the right moment can be more valuable than a full solution too early. Mistakes become much more useful when they are connected to concepts, practice, and review.

What's next

Future versions could also include spaced repetition, deeper progress analytics, teacher dashboards, and personalized study plans based on recurring misconceptions.

Built With

• biome
• css
• deepseek
• deno
• eventsource-parser
• react
• shadcn
• sse
• supabase
• tailwind
• typescript
• vite