Coaching for Teams on Agentic AI, Collaboration & Customer Sense

Your tech and non-tech teams each work with AI their own way.

A few power users pull ahead while most of the team touches the licenses you pay for once a week, or never.

Quality depends on who did the work.

The token bill grows faster than the output does.

Generic training has not fixed this. People finish a course, agree with all of it, and go back to how they worked before.

This is live coaching for teams of 5 or more.

I train your team on your real deliverables until everyone builds efficiently and uses first principles approach.

I have delivered this for 100+ PMs at Canva and 20+ at VitalAI Academy.

1. Foundations of agentic engineering : from vague vibe coding to structured development also known as Spec Engineering using agentic AI.

2. Encode first-principles such as Working-backwards spec and Bar Raiser evals that engineering can genuinely build from

3. Developing Customer-sense: using product judgment and real customer problems to improve AI-assisted work before it reaches users or leadership.

4. A guided build-and-critique session where the group applies the ideas to their own problems.

5. Live Q&A throughout the build session and after (duration and coverage depends on the size of the cohort)

Customized to your teams' needs

The curriculum is shaped to your stack and your roadmap on a scoping call. Sessions run hands-on inside the tools, and the exercises use your team's actual backlog.

Most people stick to one thing. At 31, I started on my own to experiment with multiple ventures - from teaching professionals to build better with AI to running a speciality coffee business.

This is possible because of a rather uncommon mix of experiences I packed in just over a decade across Amazon, SaaS and a top 10 MBA.

📈 Trained 15,700+ product professionals through a top-ranked free lesson on Maven

🏢 Paid cohorts have drawn 5 star reviews from people at SAP SE, Warner Bros. Discovery, PwCe

What participants say:

💬 "It's hours of jam packed practical information and advice that can be helpful at any level of AI fluency." Frank, Senior Product Manager, Big Cartel

💬 "Getting started with SPEC engineering has been much easier thanks to this workshop. I'm not a technical person, and the way they guided us step by step was a huge help." Noe, Product Manager, Wizeline

💬 "Great workshop, no fluff and provides a practical framework to build AI pm artifacts." Rushit, Enterprise Architect

Explore more resources from me.
Follow me on LinkedIn.

A spec standard engineering can build from- Working-backwards specs written and reviewed live, on a real item from your roadmap.

An eval bar that defines good before anything ships- Your team writes eval criteria and runs a bar-raiser pass, so AI output gets checked the same way no matter who built it.

A working agent per person- Everyone builds a Claude Code skill or plugin that automates part of their actual job, and leaves knowing how to extend it.

A token-spend playbook- Tight specs and acceptance criteria cut the exploratory churn that burns your AI budget.

Why this matters: teams today drain cash on AI tools but are not equipped with bespoke AI training or frameworks to use across the board. Recorded trainings become a checklist to tick off and the real gains from AI investment are missed. You do not have to take my word for it. The public versions of this material are already on the school, taught and reviewed in the open.