Your Team Talks About AI.
Ours Ships It.

A live build sprint where your product team moves from ChatGPT to building real AI products. RAG pipelines, prototypes, custom workflows. All in 8 hours.

“If you’re looking for the promised AI productivity gains everyone’s talking about, start here.”
— Lenny Rachitsky

Your Team Is Using AI Wrong

The gap between teams who treat AI as a utility and those who build it into their operating rhythm has never been wider. “Most will never catch up.”Kevin Roose, New York Times

Your team doesn’t have to be one of them.

Most teams are stuck at Level 1 or 2 of AI Adoption. Leadership buys the licenses, maybe runs a lunch-and-learn. Five percent figure it out on their own. Everyone else is waiting for someone to bring them along.

The companies moving up to Level 3–5 are asking a different question: “How do we bring everyone to the starting line together?”

Level Description What It Looks Like
1 Tool Adoption Leadership buys seats. 10% adoption.
2 Individual Builders 2–3 people figure it out. Everyone else waits.
3 Team Sport 10% → 90% adoption. Shared workflows.
4 AI Product Sense Team can evaluate AI vendors, prototype fast.
5 Product Influence AI becomes a competitive advantage.

This build sprint moves your team from Level 1–2 to Level 3–5 in 8 hours.

CASE STUDY

Zapier Turned Context Engineering Into a Team Sport

In December 2025, 65+ people across product, design, engineering, and sales engineering joined a build sprint.

Before

  • 39% had a copilot
  • Saving 3–5 hours a week
  • Half viewed AI as a tool for discrete tasks

After

  • 96% had an operating system in Cursor connected to company data
  • Projected savings of 9–15 hours a week
  • 80% now view AI as a real partner they trust
“Not all of us are AI Superstars. Investing in workshops like this, offering mentorships, and step-by-step how-tos are the best way to level up your employees who want to advance (but need a bit more hand-holding).”
“Having Glean and Databricks assistants that are trained on our data tables and can write queries is HUGE. Literally saved me hours. And hours the data team would have spent helping me.”

8 Hours. Real Products. Not Slides.

01

Personal AI OS

Build a task management system integrated with Obsidian that uses AI agents to extract action items from meetings, organize notes, and automate your daily workflow.

02

Custom AI Workflows

Design and implement RAG pipelines that let your agents query your own documents, codebases, and knowledge bases with precision and context awareness.

03

Live Prototypes

Learn the Wizard of Oz prototyping methodology to spin up convincing AI product demos in minutes, not weeks. Validate ideas before writing production code.

04

AI Product X-Ray Vision

Deconstruct any AI product to understand its underlying architecture. Distinguish genuine innovation from marketing language and ask probing technical questions.

36 Lessons Across 2 Days

Day 1

Foundations & Your Personal OS

4 hours
  • Cursor and Claude Code fundamentals
  • System prompts and agent instruction files
  • Context engineering and few-shot prompting
  • Obsidian integration and markdown workflows
  • Shell commands and git workflows
  • Building your Personal AI OS
  • Task extraction and automation pipelines
  • Hands-on collaborative building sessions
Day 2

Advanced AI Architecture

4 hours
  • RAG pipelines (with and without indexing)
  • MCP server design and implementation
  • Multi-agent systems and sub-agents
  • Browser tools and API integration
  • Wizard of Oz prototyping methodology
  • LLM optimization and evaluation
  • AI product deconstruction frameworks
  • Live prototype building workshop

Plus 30-minute breaks and optional office hours before each session. Zero homework—all building happens live.

Built for Leaders, Not Just Engineers

Product Leaders

Your team keeps talking about AI strategy. What they need is to build working prototypes in real time. This gives them that skill.

C-Suite Executives

You’re evaluating AI vendors. Half of them are vaporware. This teaches your team to spot the difference and ask questions that cut through the hype.

Non-Technical PMs

You need to prototype AI ideas fast enough to validate them before engineering spends a sprint building the wrong thing. This shows you how.

Teams in AI Transformation

You bought the seats. A handful of people figured it out. Everyone else is stuck. This gets everyone building at the same time.

Your Instructors

Aman Khan

Aman Khan

Head of Product, Arize AI

Led the deeplearning.ai course on evaluating AI agents. Featured in Lenny's Newsletter. Previously at Apple, Spotify, Cruise, and Zipline. UC Berkeley alum.

Apple Spotify Cruise Zipline
“Watching Aman build an AI agent from scratch in Cursor while explaining evals, observability, and PM-eng collaboration was eye-opening. He doesn’t just talk about AI PM skills—he shows you exactly how to develop them.” — Aakash Gupta
Tal Raviv

Tal Raviv

Early PM at Patreon, Riverside, Wix

Trained 20,000+ product people. Guest on Lenny's Podcast. Featured 3x in Lenny's Newsletter. Built products at Patreon, Riverside, Wix, AppsFlyer, and DuckDuckGo.

Patreon Riverside Wix DuckDuckGo
“I believe the future of product management looks like Tal Raviv.” — Lenny Rachitsky

What Past Participants Say

“This course’s hands-on approach uniquely combines learning with building: as you practice agentic AI concepts like MCP and memory systems, you construct your own personal assistant. You follow instructions while creating something personalized—and you continuously refine your assistant to make it increasingly useful.”
Samit Chaudhuri Head of Engineering, AWS
“Exactly as advertised and what I was looking for. 10/10 would recommend and will use on a daily basis moving forward. Easy to follow and apply, even for less-technical PMs.”
Chris DiStasio Principal Product Manager, Elastic
“As a heavy ChatGPT Projects user, this course helped me see how Cursor and Claude Code give you so many more customization options as well as unique capabilities extensible to anything I may want to do, from personal organization to building my own product. That insight, plus the practical experience, make taking the course such a win.”
Alejandro Savransky Senior Product Manager, Mozilla
“This was a great course for PMs who want to go beyond surface-level AI usage. I got started building a personal OS that already feels like a durable foundation—something I can keep iterating on and actually use in my work, not just a demo.”
Shikha Sharma Product Manager, Visa

Ready to Build AI Product Sense?

Book a Build Sprint

Schedule a 30-minute call to discuss your team’s AI transformation goals.

Questions? Email amanaiproduct@gmail.com