How to use Claude Code to build a board room of world class Product Mentors

🧠 Steve Jobs, Demis Hassabis, Ray Dalio, Sheryl Sandberg and more. Build your own Product board of world-class mentors. Knowledge Series #98

🔒The Knowledge Series is available for paid subscribers. Get full ongoing access to 90+ technical explainers and AI guide and tutorials to grow your technical knowledge at work. Multiple new guides are added every month.

---

A startup founder recently revealed that they use a custom built AI version of Steve Jobs to help guide their decision making. In this case, her GPT was for her brand consultancy and whenever she’s stuck on a strategic decision or needs a creative idea, she refers to her Steve Jobs clone. She’s trained it on product launch transcripts and strategy talks so that when she needs creative input it will critique her ideas and help them refine them towards a “10 out of 10” concept.

A brand consultancy is a little different to a tech company, but what if you could build an entire team of world class product thinkers and refer to them to help you to assess your designs, make decisions or craft your strategy?

It’s slightly dystopian, but Claude Code is now increasingly being adopted by non‑engineers, and its ability to reference markdown persona files, as well as capabilities like Agents and Skills, makes this idea possible in 2026 for product teams who want to occasionally lean on some world‑class expertise for decision making - just out of curiosity, if nothing else.

In this Knowledge Series, we’re going to go step by step through the process of building your own board room of world class product mentors that you can chat to across each of their specialist areas including:

• Strategy and vision setting

• AI and product sense

• Decision making and frameworks

• Design and craft

• Communication and storytelling

• And more

We’ll cover how to get Claude Code set up so that you can create your mentor files - and we’ll transform the mentor system into a usable front end so that you’re not stuck interacting through the command line. Plus, we’ll take a look at how you can connect your mentor system to third party tools like Linear, Figma, Notion and others so that your mentors’ decision making is grounded in real world data.

If you’re curious about how non-engineers can use Claude Code for use cases at work outside of Code then this Knowledge Series should help.

---

The Knowledge Series

How this Knowledge Series is structured

This Knowledge Series is broken down into 4 core parts:

1. Choose your product mentors

2. Build your mentor system with unique skill sets in Claude Code

3. Turn your mentors into an interactive chat application you can use whenever you need to

4. How to connect external tools like Linear, Figma, Jira and others to your system

Example of a Product Mentor boardroom with leaders from Deepmind, Stripe, Meta, Apple and more

First, we’ll talk through some of the principles and frameworks you can use to choose your mentors. For the best results, choose mentors that you admire who have specialist expertise in whatever areas you think are beneficial for you. These don’t even need to be people in tech but in our examples we’re going to stick mostly to tech execs.

Once you’ve chosen your mentors, we’ll go through some of the prompt techniques you can use to create a detailed description of their thought processes that will form the basis of their capabilities.

Finally, we’ll turn our mentor system into a simple, functional app with a front end that you can use whenever you need some guidance - as well as ideas on how to build useful AI Agents and Skills that will take your mentor system to the next level.

Getting set up

For this Knowledge Series, we’re going to use a mixture of Claude through the web and Claude Code in Cursor. You can read the full instructions on how to get Claude Code set up in the previous Knowledge Series here but once you’ve installed Claude Code you’ll then be able to reference it in Cursor’s own terminal by typing “claude”:

We won’t need to use Claude Code just yet - instead we’ll first build our product mentors using Claude for the web and then store these to use in our Mentor system later.

Choose your product mentors

Your mentors can be anyone you like. For this Knowledge Series, we’ll choose our mentors based on 19 different skill dimensions including:

• Strategy and vision

• Discovery and research

• Execution and delivery

• Design and craft

• People and communication

• Data and measurement

• AI Product Sense