AI math tutor with worked examples
Walk learners through problems step-by-step, ask why each step works, and give immediate feedback on misconceptions, grounded in your textbook, not the open internet.
Coming soon
Build an AI tutor for any subject with Claude or ChatGPT, ready for any LMS.
Build an AI tutor that explains concepts, asks Socratic questions, and adapts to each learner. Describe the topic and the learning objectives in plain language, your coding agent produces a SCORM-ready tutor for any LMS.
What is it?
What is an AI tutor?
A patient, always-available AI tutor that teaches a specific topic through dialogue, checks understanding with targeted questions, and adapts its explanations to the learner's level, all graded against your learning objectives.
-
Conversational tutoring on any topic you describe, from cell biology to SQL joins to GDPR Article 6 -
Adapts explanations and difficulty to each learner's responses, not a static script -
Checks understanding with Socratic questions and graded mini-quizzes throughout the session -
Cites the source material you provide, so answers stay grounded in your curriculum, not the open web -
Exports as a SCORM 1.2 package for Cornerstone, Moodle, Canvas, TalentLMS, Docebo, Brightspace, and every LMS -
Same bundle runs standalone in any browser for self-study without an LMS
How it works
How to build an AI tutor
in four steps.
From a topic description to a graded, SCORM-ready tutoring session. No scripting language, no LMS plumbing, no model wrangling.
01
Describe the topic and source material
Paste the chapter, policy doc, API reference, or lab manual. The tutor grounds every answer in what you provide so it doesn't hallucinate around your curriculum.
02
Define the learning objectives
Name what the learner should be able to do, explain, or apply by the end of the session. The tutor checks each one with targeted questions before moving on.
03
Tune the tutor in chat
"Be more Socratic." "Drill harder on edge cases." "Use worked examples." Iterate by talking with your coding agent, not by editing prompts.
04
Export as a SCORM 1.2 package
One command produces a SCORM 1.2 zip. Upload to any SCORM LMS. Objective-level mastery scores post back to the gradebook automatically.
Use cases by role
AI tutoring
for every subject and audience.
The same toolchain powers AI math tutoring, language practice, onboarding assistants, compliance refreshers, and corporate upskilling. Describe the topic, the tutor adapts.
Conversational language partner
Practice a target language with an AI tutor that corrects gently, varies difficulty, and explains grammar on demand. Transcripts grade vocabulary and structure use.
New-hire onboarding tutor
Turn your handbook and SOPs into a one-on-one onboarding tutor that quizzes new hires on policies and posts scores back to your LMS.
Compliance refresher tutor
Replace 30-slide annual training with a 10-minute adaptive tutor that drills exactly the policies each learner is shaky on, then certifies completion in your LMS.
Developer onboarding tutor
Teach your internal API, migration playbook, or incident-response runbook. The tutor asks engineers to explain and apply concepts, not just acknowledge them.
Clinical protocol tutor
Walk clinicians through diagnostic reasoning or safety protocols with targeted questions and immediate feedback, grounded in your clinical guidelines.
Objective-level mastery scoring
Every session graded against your learning objectives, turn by turn.
When the session ends, the AI scores the full conversation against your objectives. Each one comes back with a mastery rating, evidence pulled from the transcript, and a specific next-step recommendation, every score auditable against the words that produced it.
Update objectives or weights any time. Re-export the SCORM package and the next attempt grades against the new rubric.
Per-objective mastery scoring
Each learning objective is graded separately based on the learner's responses to checks for understanding. No single opaque pass/fail.
Adaptive difficulty
The tutor backs off when a learner struggles and presses harder when they're cruising. No fixed difficulty curve, no wasted time.
Grounded in your sources
Citations point back to the chapter, page, or section you provided. Learners and auditors can verify every answer.
Session complete
Your Debrief
9/10
★★★★★★★★★☆
Total 9 / 10
Explained the underlying concept in the learner's own words before applying it to a new example.
Identified at least one misconception during the session and resolved it before moving on.
Demonstrated mastery on the transfer question without re-prompting from the tutor.
Plain-English authoring
Author AI tutors in plain English. No scripting required.
Tweak the persona, swap objectives, change the teaching method, or rewrite the rubric, just describe the change in your own words. Your coding agent rebuilds the tutor and re-exports the SCORM package without you ever touching a config file. No proprietary DSL, no template format, no lock-in.
When you're happy, export as a SCORM 1.2 package for your LMS, or use the same HTML bundle standalone.
Make the tutor more Socratic on the SQL joins topic.
Updated. Tutor now asks 'what do you think happens next?' before stating any join result.
Add a mastery check on left joins with nullable foreign keys.
Added objective
left-join-nullables (weight 3). Re-grading rubric.Export as SCORM.
✓ ai-tutor.zip ready (SCORM 1.2, 14 KB)
Tutor Library
Start from real classroom and workplace
tutoring sessions.
Paste any prompt into your coding agent to get a complete AI tutoring session as a single self-contained HTML file. Adjust the topic, rubric, or teaching method in plain language.
Math
Algebra to calculus bridge
A patient AI tutor walks a learner from limits intuition to the formal derivative definition, catching algebra slips before they compound.
Prompt
Using /ai-tutor, build a 14-turn algebra-to-calculus bridge tutor. Score on limit intuition, derivative definition, and clean algebra.
Language
Spanish past-tense practice
A conversational Spanish tutor that varies difficulty as the learner attempts preterite vs imperfect, with targeted grammar reminders only when needed.
Prompt
Using /ai-tutor, build an 18-turn Spanish tutor on preterite vs imperfect. Score on correct tense, accurate aspect, and self-correction.
Onboarding
PTO policy walkthrough
A new-hire tutor that teaches the company PTO policy and confirms understanding with scenario questions, grounded in the actual handbook.
Prompt
Using /ai-tutor, build an 8-turn PTO-policy tutor for new hires. Score on accrual math, blackout dates, and approval workflow.
Engineering
Internal API onboarding
A code-aware tutor that introduces a new engineer to the team's internal HTTP API, asks them to predict response shapes, and corrects misconceptions.
Prompt
Using /ai-tutor, build a 12-turn internal-API tutor. Score on endpoint semantics, error handling, and pagination model.
Compliance
HIPAA refresher
A short, sharp HIPAA tutor that focuses on real scenarios reps actually encounter, not a 40-slide deck. Adaptive depending on what the learner already knows.
Prompt
Using /ai-tutor, build a 10-turn HIPAA refresher. Score on PHI identification, minimum-necessary rule, and disclosure exceptions.
Healthcare
Diabetic foot exam protocol
A clinical tutor that walks a learner through the diabetic foot exam, asking them to justify each step against the underlying clinical guideline.
Prompt
Using /ai-tutor, build a 16-turn diabetic-foot-exam tutor. Score on inspection, monofilament test, and risk stratification.
SCORM & LMS
SCORM 1.2 AI tutoring activities for any LMS.
Every AI tutor exports as a standards-compliant SCORM 1.2 package. Upload the zip, assign it like any other course activity, and completion plus mastery scores flow back to the gradebook automatically.
-
Single SCORM 1.2 zip, upload to your LMS, no integration work, API keys, or custom JavaScript -
Completion, score, and time-on-task post back via cmi.core.score.rawandcmi.core.lesson_status -
Tested with Cornerstone OnDemand, Moodle, Canvas LMS, TalentLMS, Docebo, Brightspace, Absorb, 360Learning, SuccessFactors, and Workday Learning -
Same self-contained HTML bundle runs standalone on a public page, an intranet, or as an embedded iframe -
Bring your own model: Claude, GPT, Gemini, or a self-hosted alternative, swappable at any time -
Grading runs inside the bundle, no middleware, no data warehouse, no analytics SDK required
FAQ
Frequently asked questions about
AI tutors.
Authoring, grading, LMS compatibility, data privacy, model choice, answered directly.
An AI tutor is an interactive learning activity where a learner has a one-on-one conversation with an AI that teaches a specific topic, checks for understanding with targeted questions, and adapts its explanations to the learner's level. The session is graded against instructor-defined learning objectives.
A generic chatbot answers whatever the user asks, from whatever knowledge it was trained on. An AI tutor is grounded in your specific source material, follows a defined set of learning objectives, checks for understanding rather than just answering, and produces a graded transcript for your LMS.
Any subject you can describe and supply source material for, math, language, science, coding, healthcare, law, compliance, sales methodology, internal policy. The tutor learns the domain from the materials you provide.
Yes. Every AI tutor exports as a SCORM 1.2 package, which works in Cornerstone, Moodle, Canvas, TalentLMS, Docebo, Brightspace, and every other SCORM-compliant LMS. Objective-level scores post back to the gradebook automatically.
The tutor checks understanding throughout the session with targeted questions tied to your learning objectives. Each objective gets a mastery score based on the learner's responses, evidence pulled from the transcript, and a specific next-step recommendation.
The tutor is grounded in source material you provide and cites that material in its responses, which dramatically reduces hallucination. You can also turn on strict mode, where the tutor refuses to answer questions outside the supplied material.
The bundles run on whichever frontier model you configure, Claude, GPT, Gemini, or a self-hosted model. The model is swappable; the rubric, persona, and scoring logic are model-agnostic.
AI Tutor is in active development. It will install via your coding agent with a single command, the same way the existing role-play skill works today. Add your email below and we'll let you know the day it ships.