What AIxPM is, and who is behind it

AIxPM is a free site for practising product management in the AI age. 61 cheatsheets, 64 scenarios, an 18-lesson curriculum, a 31-book reading shelf and a 12-question judgment check. No sign-up, no paywall, no ads.

Who built this

Karthik Elangovan. Product manager at a B2B SaaS company, shipping AI features and getting a fair number of the calls wrong along the way. This is a side project, built and maintained by one person, with help from AI tools for the drafting and a lot of hand review.

Why it exists

I always felt that if there was an organised way to practise PM skills, the way you can practise an instrument, that would be a goldmine. There was not one. So I collected the material I had bought over the years, built a study plan for myself and worked through it. It helped more than anything else I had tried.

This site is that plan, cleaned up so other people can use it. It is not solving a market problem and it is not a funnel for anything. If it is useful to you, good. If not, it was still worth it for me.

How the questions are made and judged

Every question starts from one of the 64 practice scenarios, each of which has a written senior-PM answer. The scenario is expanded into multiple-choice questions with a language model, where the wrong options are the things PMs actually say in those meetings.

A second model then judges each question for a single defensible answer and for distractors that are plausible rather than silly. Roughly one question in eight is rejected or flagged at that step. The ones that pass are reviewed by hand before they go live. Rejected questions stay in the database with their verdict so the bank can be audited.

• Questions are tagged to one of six dimensions: metrics, evals, discovery, strategy, influence and AI judgment.
• The avatar check samples from the approved bank only, with the same difficulty mix for everyone on a first attempt.
• Every wrong option carries a short rationale, so a miss tells you what the trap was, not just that you fell for it.

Where the material comes from

The cheatsheets distil public writing and talks from people who do this work: Lenny Rachitsky's newsletter and podcast, Lewis C. Lin's interview material, the books on the reading shelf, and a handful of practitioner essays. Each cheatsheet links its sources at the bottom. Nothing here is a copy of a paid course.

The book takeaways are written for AI product managers specifically. They are not summaries of the books. Read the book.

What is stored

Nothing, unless you ask. Progress lives in your browser. If you choose to save it to another device, it is stored against your email and a four-digit code you pick, and it holds quiz results only. Avatar results feed an anonymous percentile with no identity attached. There is no tracking beyond standard page analytics.

Report a bad question or a wrong fact

If a question has two defensible answers, feels unfair, or is simply wrong, flag it with the button under the question. Flags land on an admin review list and the question is pulled or fixed. For a mistake in a cheatsheet, use the feedback box at the bottom of the page. Both are read.