How to Use PrepGraph

A guide to the product — what it is, what it contains, and how to get the most out of it.

Marjuk · 2026-04-06

The philosophy behind the product

Most people pursuing PM and DS roles consume a lot. Articles, newsletters, YouTube videos, courses. The problem is not access to information; it is retention. You read something on Monday and by Thursday it is gone.

PrepGraph is built on a simple premise: knowing something and having read about something are not the same thing. The product is designed to close that gap, to move concepts from "I have seen this" to "I genuinely understand this and can use it." This is what the Workshop does. You can track concepts by topic and return to them whenever you wish. It is powered by Claude Opus 4.6.

The way it does that:

• Structured practice. Not passive reading. You answer questions, get AI feedback, and earn your mastery level.
• Active tracking. Every concept you capture has a depth: Surface, Understood, or Mastered. You can see exactly where you stand.
• A knowledge base that compounds. Everything you practice, every note you write, every article you extract sits in one place and connects over time.

Interview readiness is a side effect of this. When you genuinely know the material, interviews take care of themselves. That is not the goal. It is the outcome.

[architecture-diagram]

Getting started

When you first sign up, PrepGraph asks you two things:

1. Your track — Product Management or Data Science. This determines which curated concepts get seeded into your graph.
2. Your timeline — How far out is your next interview or goal? This determines how many concepts get added on day one.

After that, your workspace is ready. You have a set of concepts organised by topic, each mapped from real interview loops and learning curricula. The system is pre-loaded; you do not start from a blank page.

Arena

The Arena is where you practice.

It is built around the curated syllabus: 94+ concepts across PM and DS tracks, each with a difficulty rating, company tags (which companies test this), and a specific interview question.

How a session works:

1. Pick a topic area from the tabs at the top: Metrics, Product Sense, Machine Learning, Experimentation, and so on.
2. Open a concept. You will see what companies test it, what a great answer covers, and the interview question.
3. Type or speak your answer. There is a microphone button for voice dictation, useful if you want to practice out loud as you would in a real interview.
4. Hit Get AI feedback. The AI evaluates your answer against the criteria and returns a rating: Needs Work, On Track, Strong, or Outstanding. It tells you what you got right, what was missing, and shows a model answer.
5. Alternatively, hit Model answer first, useful when you want to understand the concept before attempting it yourself.

Hints are available if you are stuck. They reveal one at a time; the idea is to nudge you in the right direction without giving the answer away.

Save to notes appears after feedback. One click and the AI's assessment gets saved permanently to that concept's notes. Every session makes your knowledge base richer.

At the top of the Arena you can see your Interview Ready %: the percentage of syllabus concepts you have mastered. This number is the honest answer to "am I ready?"

Ask your knowledge base sits at the bottom of the Arena. Ask anything across your entire graph. "What connects North Star Metric to experimentation design?" or "What are my weakest areas in Metrics?" The AI answers from your own captured knowledge, not generic internet content.

Workshop

The Workshop is your personal knowledge base.

Think of it as the map of everything you know. The Arena is where you practice. The Workshop is where it lives.

What is in it:

• Every concept you have added, organised by topic
• The mastery depth for each concept (Surface, Understood, Mastered)
• Your notes on each concept
• A review queue: concepts due for spaced repetition review based on how long it has been since you last touched them
• Your rank and neuron count, which reflect how much you have learned overall

Adding concepts:

You can add concepts in three ways:

1. Manually — hit "Add concept", give it a name, assign it to a topic, set your current depth.
2. From the Arena — when you add a syllabus concept to your study set, it appears in the Workshop automatically.
3. From source — paste any article, paper, or transcript and hit "Extract concepts". The AI reads the text, identifies 3 to 8 PM or DS concepts from it, summarises what the text says about each one, and suggests which topic they belong to. You pick which ones to add.

This third method is the most powerful. It means any content you read can feed directly into your knowledge base. You are not just consuming; you are building.

Ask your knowledge base is also available in the Workshop. Same feature, same power. The more you have added, the more useful it becomes.

The Graph

The Graph view shows how your concepts connect to each other visually. As you build out your knowledge base, patterns emerge: clusters of related ideas, gaps between topic areas, concepts that act as bridges between domains.

It is not the primary way to interact with the product, but it is worth visiting occasionally. Sometimes seeing your knowledge laid out spatially reveals something a list cannot.

A note on mastery depth

PrepGraph uses three levels: Surface, Understood, and Mastered.

• Surface: you have encountered this concept. You know it exists. You could not explain it confidently.
• Understood: you can explain it in your own words. You get the idea.
• Mastered: you can apply it, teach it, and answer questions about it under pressure.

These are not self-reported. They are earned through practice. The AI feedback tells you where you actually stand. Move a concept to Mastered when the feedback consistently says Strong or Outstanding, not before.

The honest gap between what you think you know and what you can demonstrate is exactly what PrepGraph is designed to close.