StudyRescue AI

StudyRescue AI — Turn limited study time into smarter exam preparation.

Comment

The Problem

Students rarely struggle because they have no study plan. They struggle because their plan becomes outdated when they get questions wrong, discover weak concepts, or simply run out of time.

Traditional study planners treat every topic equally and assume the original schedule will work.

StudyRescueAI takes a different approach.

The Idea

StudyRescueAI is an adaptive learning system designed for students who have limited time before an exam.

Instead of asking:

"What should I study today?"

StudyRescue asks:

"Given what I know right now, what should I study next?"

The system analyzes assessment performance, identifies high-risk topics and concepts, and allocates limited study time toward the areas where the student needs it most.

How It Works

The experience follows an adaptive loop:

Diagnose → Prioritize → Rescue → Adapt → Recalculate

A student first completes an assessment. Their answers create evidence about their current mastery.

When they start a Rescue Session, StudyRescue targets weak concepts instead of randomly selecting questions.

If the student answers incorrectly, the system identifies the associated misconception, provides a quick correction, and uses that evidence to select the next appropriate practice question.

Difficulty can also adapt based on performance.

After the Rescue Session, the system combines the new evidence with the student's previous performance and recalculates mastery, risk, priorities, and study-time allocation.

This means the plan can actually change because the student learned something.

What Makes It Different

StudyRescue is not just a calendar or a chatbot wrapped in a study dashboard.

The core of the project is the adaptive learning-state engine:

  • Concept-level performance tracking
  • Risk-based prioritization
  • Targeted misconception follow-up
  • Adaptive question difficulty
  • Question history and repetition prevention
  • Mastery updates from new evidence
  • Dynamic study-time allocation
  • Rescue-plan recalculation when the student falls behind

The key idea is simple:

Don't study everything. Study what you need most.

How We Built It

The application was built with React, TypeScript, Vite, Zustand, Framer Motion, and React Router, and deployed on Cloudflare Pages.

We focused heavily on making the adaptive behavior reliable rather than building a static prototype. The Rescue flow was tested around question selection, misconception follow-up, difficulty adaptation, mastery updates, question history, and session completion.

What We Learned

The biggest lesson was that adaptive learning is not just about generating questions.

The difficult part is maintaining a reliable learning state: knowing what the student answered, what concept they struggled with, what evidence was just collected, and how that evidence should change the next action.

That became the central engineering challenge of StudyRescueAI.

What's Next

We would expand StudyRescue with broader subject coverage, spaced repetition, long-term learning analytics, teacher/mentor dashboards, and real-world learning-outcome validation with students.

StudyRescueAI is built around one principle:

When time is limited, every answer should help decide what happens next.

Built With

Share this project:

Updates