AI Problem-Solving Framework for Students

How it works

From problem description to guided problem-solving using Polya's 4 steps.

Describe the problem, any type, any subject

Priya is an 8th grader facing a math word problem: 'A farmer has 100 feet of fencing and wants to enclose the largest possible rectangular area. What dimensions should the rectangle have?' Her teacher types the problem into the form. The AI can identify this as an optimization problem requiring both algebraic reasoning and spatial visualization, and can recognize that the problem type often confuses students because the constraint (perimeter) and the goal (maximum area) are easy to conflate.

Phase 1, Understand the Problem: guided clarification questions

The AI generates questions for Phase 1 of the Polya framework: What is the unknown? What are the given conditions? Is there enough information to solve this, or is something missing? Can you draw a diagram or restate the problem in your own words? For Priya's fencing problem, the questions include: What does 'largest possible area' mean as a mathematical quantity? What formula connects the fencing (perimeter) and the dimensions you are looking for?

Phase 2 and 3, Devise and carry out a plan with strategy suggestions

For Phase 2, the template lists five strategies: draw a picture, make a table, look for a pattern, try a simpler case, or work backwards. For a fencing problem, the AI can suggest starting with a table of values (trying several dimension combinations that satisfy the perimeter constraint and comparing their areas) before moving to an algebraic generalization. This builds intuition before formalism.

Phase 4, Look Back: metacognitive reflection prompts

After Priya reaches a solution, the Look Back questions the AI generates ask: Does the answer make sense? What would happen if the total fencing changed (does your method still work? Can you find this result a different way? Could you use this same approach on a different problem? The Look Back phase is the one students most often skip, and it is where the transferable problem-solving habit actually forms) connecting this specific solution to a general method.

The non-routine problem gap

Most classroom math instruction teaches students to execute known algorithms on problems that look like the worked examples. Students become fluent with routine problems but struggle with non-routine problems, where the solution method is not immediately obvious and where the student must select and sequence strategies without a template to follow.

Standardized tests, college coursework, and workplace challenges are dominated by non-routine problems. The Problem-Solving Framework gives students a printed routine for practising metacognitive habits on problems they have not seen before.

What the problem-solving framework includes

Polya's 4 phases. Strategy suggestions. Metacognitive reflection. Transferable habits.

Polya's 4-step framework built in

George Polya's problem-solving framework (Understand, Devise a Plan, Carry Out the Plan, Look Back) is one of the most researched and validated frameworks for mathematical and scientific problem solving. The template structures every guide around these four phases, with questions at each step. Using the framework often can help students notice their own problem-solving process.

Works for math, science, and open-ended problems

The framework is not limited to algebra or geometry. It works for math word problems (optimization, rates, combinatorics), science problems (experimental design, data interpretation, cause-and-effect reasoning), and open-ended design challenges (engineering design challenges, project-based learning scenarios, case studies). The guided questions the AI generates adapt to the problem type the teacher types.

Problem-solving strategy suggestions

At the Make a Plan step, the template lists five strategies: draw a picture, make a table, look for a pattern, try a simpler case, and work backwards. The AI's suggestions are problem-specific, as not all strategies are equally relevant for every problem. The AI selects two useful strategies for the specific problem the teacher typed.

Metacognitive reflection prompts

The Look Back phase generates questions that develop the metacognitive habits that transfer problem-solving ability across contexts: Does the answer make sense in context? Can you verify the solution a different way? What would change if one constraint were different? How is this problem similar to problems you have solved before? Asking these questions as a habit can help students with the problem solving that standardized tests, college coursework, and careers require.

Non-routine problem support

The tool is specifically designed for non-routine problems (problems where the solution method is not immediately obvious and where applying a memorized algorithm is insufficient. Routine problems (solving a quadratic, finding a derivative) are better handled by the AI Equation Solver. The Problem-Solving Framework is for problems that require selecting and sequencing strategies) which is exactly the problem type that appears on standardized math exams and in real-world contexts.

Builds transferable problem-solving habits

The goal of the tool is not to help a student solve one specific problem, it is to help them internalize a problem-solving process that works across problem types, subjects, and contexts. With regular use across problem types, students can start applying the phases automatically: identifying unknowns and conditions before attempting a solution, planning before computing, and verifying answers before submitting.

Who uses the problem-solving framework

Teachers preparing students for AP or IB exams give them the guide for free-response problems and extended response questions where the solution path is not predetermined, exactly the format that appears on AP exams and IB Internal Assessments.

Teachers in project-based learning environments give students the guide for the problem-definition and solution-design phases of PBL projects, where the structure helps them move from an open-ended challenge to a specific, actionable plan.

Teachers teaching problem-solving as a skill use the tool to make the four phases explicit during class, showing students that problem-solving is a learnable process, not an innate ability, by working through the guided questions together on a projected problem.

Teachers of students who freeze on unfamiliar problems use the guide as a starting point, the Understand the Problem questions give them a structured entry point that prevents the blank-page paralysis that unfamiliar non-routine problems often trigger.