Skip to content

How it works.

PreMedulla runs the diagnostic and planning work so you can focus on studying. Here's what that looks like from your first login.

The diagnostic finds your weak areas in minutes.

You start with a self-scored diagnostic. It walks you through a set of questions, you mark what you got right and wrong, and PreMedulla maps the results across all four sections.

The output is a personalized study plan: the sections and topics where you're losing the most points, ranked by impact. You don't have to figure out what to work on. The plan tells you.

This runs entirely in the app with no AI and no subscription. You see your plan the moment you finish.

The right chapters and concept checks queue up for you.

For each weak area, PreMedulla queues the textbook chapters that cover it, followed by concept checks to confirm you actually understood it. You don't hunt for what to read. You open the app and it's there.

Concept checks are short and self-scored. Pass one and the topic moves toward mastered. Miss it and the plan keeps it on your radar.

Flashcards round it out. The deck is built from your weak areas, not a generic card set.

Mastery tracks across all four sections automatically.

Every concept check and practice attempt updates your mastery picture. The dashboard shows each section with the topics inside it, so you can see at a glance what's improving and what still needs work.

You don't have to keep a spreadsheet or remember what you last studied. The app holds that state for you. That's the medulla part: the tracking running underneath, so the work running on top is the work that actually moves your score.

The plan updates as your mastery changes. Concepts you've solidified drop down the priority list. Concepts that keep coming up stay near the top.

Practice the way that actually builds your reasoning.

When you practice passages, you commit your reasoning before the choices unlock. You read the question, write down what you think the answer is and why, then reveal the choices and pick one. Then you check your pick against the answer and the written explanation.

This is predict-mode practice. It forces you to make your reasoning visible before you see the options, which stops you from reverse-engineering an answer after the fact. Self-checking against the explanation tells you where your reasoning diverged from the passage.

Paid tiers add AI grading of your written reasoning, which goes deeper on the specific move you made and whether it held up. That's coming. The free practice method works without it.

What the tutor covers today.

Honest about scope. The tutor is complete for the sections it covers, not half-built.

Available now

CARS · Psych/Soc

Diagnostic, personalized study plan, chapters, concept checks, flashcards, and mastery tracking for both sections. Predict-mode practice with self-check against the written explanation. AI reasoning grading is a paid upgrade (coming).

Coming next

Bio/Biochem · Chem/Phys

Same engine, same plan-and-track loop, different content. They ship when paying users say they're the priority, not on a marketing calendar.

What the tutor doesn't do.

PreMedulla doesn't reteach biology, organic chemistry, or the rest. When your profile shows a real content hole, it points you to the resources that already do it well. The reasoning and pattern layer is the part we own; the content layer already exists and we point at it.

  • Sketchy
  • Khan Academy
  • AnKing
  • AK Lectures

See it in action.

Click through the real app screens with a fictional sample student. Study plan, predict-mode practice, calendar, flashcards, textbook, and analytics.

Try the interactive demo

Sample data only. No account needed.

Get started free.

Sign up and build your personalized study plan in minutes. The diagnostic, plan, chapters, concept checks, flashcards, and mastery tracking are all free. No card needed.

Get started free

Create your account and your plan is ready in minutes.

Get started.

Sign up free. We'll send your first diagnostic to your inbox.

No card. No catch. One free diagnostic.

Questions about how it works.

What does "free" actually include?

The free tier includes everything you need to build and follow a structured plan: the self-scored diagnostic, your personalized study plan, the textbook chapters and concept checks queued for your weak areas, flashcards, and mastery tracking across all four sections. No AI runs on the free tier. AI reasoning grading on your practice answers is a paid feature, coming with the first paid tier.

What does predict-mode practice mean?

Before the answer choices appear, you write down your answer and reasoning in your own words. Then the choices unlock, you pick one, and you compare your pick against the correct answer and the written explanation. This forces you to commit your reasoning before you can see the options, which stops you from reverse-engineering a choice after the fact. The AI grading of your written reasoning is the paid upgrade on top of this method.

What happens to my data?

Your practice attempts and mastery data are yours. They live in PreMedulla and don't get sold, shared, or used to train anyone else's model. If you cancel, you can export or delete your data.

Where does AI actually run?

Two places, both paid. First: reasoning grading on practice answers, where the AI checks your written reasoning against the passage and the correct explanation. Second: the CARS tutor, which can walk through a passage with you. The diagnostic, plan, chapters, concept checks, flashcards, and mastery tracking are all deterministic, no AI. That's what keeps the free tier genuinely free to run.

Can I see the trap taxonomy?

Yes. The CARS error taxonomy is visible in the practice debrief. We don't gate it. It's the shared language you and the grader use to talk about a miss, so hiding it would defeat the point.

How is this different from just using ChatGPT?

ChatGPT answers questions you ask it. PreMedulla finds your weak areas, builds your study plan, queues your chapters and practice, and tracks your mastery over time. It forgets nothing between sessions. That persistent picture of where you stand is the thing a general chatbot can't replicate.