Features
MathRoutine combines higher-quality word problems, controlled AI scaffolding, equation-chain checking, Pro FRQ rubric feedback, and weakness memory across Grade 1-12 so practice becomes more than another worksheet.
Practice by grade, topic, and difficulty with problems that surface hidden quantities, comparison logic, reverse reasoning, and multi-step setup.
Students try first, then use one support tool at a time: key information, guided steps, simplified English, vocabulary, or translation.
Final answers and equation chains are checked separately so feedback can distinguish wrong setup, calculation slips, and wording confusion.
Pro written-response practice checks claim, setup, justification, notation, and final conclusion so upper-grade students can practice AP-style reasoning, not just final answers.
Translate and simplify the problem wording without replacing the math reasoning the student still needs to do.
Repeated patterns are treated as learning signals, so a missed structure can return later as focused comeback practice.
Reports explain what changed: accuracy, pacing, weak topics, reasoning demands, and the next better practice target.
Account setup avoids unnecessary required fields, keeps school optional, and makes privacy, AI-feedback limits, billing, and deletion paths visible before families subscribe.
Example
A paid student should not see only easy arithmetic dressed up as a story. Strong word problems should force them to identify the model, ignore distracting information, and explain the equation.
Grade 10 exponential model
A server begins with 600 requests and triples each year. The limit is 16,800 requests. What is the greatest whole number of years the server can grow without reaching the limit?
Highlight key info
Start at 600. Multiply by 3 each year. Stay below 16,800. The phrase "without reaching" means strict inequality, not equal to the limit.
Strong equation chain
600 * 3^n < 16800, so 3^n < 28. Since 3^3 = 27 and 3^4 = 81, the greatest whole number is n = 3.
Misconception-specific feedback
If a student answers 4, the issue is not arithmetic. The student treated "without reaching the limit" as if reaching or passing the limit were allowed. The next practice should focus on strict inequality language.
What MathRoutine Diagnoses
Families should not pay for another answer checker. MathRoutine is designed to separate the miss into a learning decision: setup, calculation, vocabulary, hidden quantity, final-question confusion, or written-reasoning quality.
Setup miss
The student chose the wrong relationship before calculation started.
Calculation slip
The model is right, but arithmetic, simplification, or units failed.
Vocabulary blocker
The math is accessible, but wording such as total, more than, remaining, or rate hides the operation.
Hidden quantity
The student used the visible number and missed the amount implied by the story.
Final-question confusion
The work answers an intermediate quantity instead of the actual question.
FRQ reasoning gap
The response needs a claim, evidence, notation, and justification to earn rubric credit.
Upper-Grade Proof Preview
Grade 11-12 practice should not collapse into quick algebra. MathRoutine Pro includes FRQ-style tasks where students must state an interval, justify endpoint behavior, use notation correctly, and explain why a conclusion follows.
Calculus BC rubric-style prompt
A power series centered at x = 1 has ratio-test convergence when |x - 1| < 5. At x = -4, the endpoint series is alternating and converges conditionally. At x = 6, the endpoint series is harmonic. State the full interval of convergence and justify both endpoint decisions.
Credit point 1
States the open interval from the ratio test before checking endpoints.
Credit point 2
Includes x = -4 because conditional convergence still counts.
Credit point 3
Excludes x = 6 and explains that harmonic divergence overrides the open-radius result.
What the grader should catch
A final interval alone is not enough. The response needs endpoint evidence, correct bracket choice, and a reasoned distinction between conditional convergence and divergence.
Feedback Intelligence
MathRoutine checks the final answer and the equation chain separately. On Pro FRQ practice, it also checks whether the written response earns rubric credit for setup, justification, notation, and conclusion. That lets the feedback distinguish calculation slips, weak proof, wrong setup, and repeated misconception patterns.
Correct setup, wrong answer
Keep the equation and repair arithmetic, units, or simplification.
Correct answer, weak equation
Ask for a clearer model so the student can prove the reasoning.
Repeated related miss
Flag misconception memory and schedule focused review.
FRQ response needs more proof
Award partial credit signals for setup, justification, notation, and conclusion instead of treating the answer as simply right or wrong.
Higher-order blocker
Track hidden quantity, reverse reasoning, comparison logic, or multi-constraint setup.