AI Agent Builder Academy Founders & CEOs (EN) Founders & CEOs (DE) Marketing Leaders Product Leaders
Product Edition

Get your product team building AI agents that move the metric

Four weeks, live, for product leaders and their teams. Agents that rank what to build next, find why users leave, and prove which change lifted conversion. You’ll build the ones that matter to you.

Four weeks, live · from £1,000 + VAT

Participants of previous AI Agent Builder Academy programmes include senior leaders from
Bolt
EY
Angi
Trip.com
Egon Zehnder
Mollie
Checkatrade
Blue Light Card
Ginestié Avocats
Sprints
AB Capital
Founders Keepers
Gino Legal Tech
Banco de Crédito BCP
Mantra Investment Partners
DoorFeed
Pix.City
Matrix One
JBM
The End TV
Albertine
Ocean Born Foundation
Bold Print Studios
Sproot Baby
What’s included
1

Live workshops

A live group session every week: a Product use case built live, Show & Tell, and Q&A with the room.

2

1-1 coaching

Thirty minutes a week, one to one with Raj. Bring your strategy, or bring the agent you are stuck on.

3

Recorded modules

A new module every Friday, yours to watch in your own time and to keep after the programme ends.

4

A cohort of product teams

You learn by building your own, and by seeing what other product leaders are building: what is working, what failed, and why.

About 2 to 3 hours a week, plus homework, with optional office hours on Monday and Wednesday. Built to fit around a real job.

Your host

Taught by a product leader who has owned the outcomes.

Rajinder Ashan
Rajinder AshanPartner — AI for Product

I have been in product since 2013, doing the work before the industry had a name for it. Growth, internationalisation, B2C, B2B, back-office, and the teams behind all of it, at UK and European businesses including Trainline and CarNext.

Until early 2026, building agents was developer territory. Then the tools got good enough for the rest of us. That is the whole opportunity.

In four weeks I will teach you to build your own. The measure is not how much you ship. It is whether the numbers move.

Every workshop and every 1-1, you learn from me directly.

The four-week curriculum

From AI user to AI builder, in four weeks.

Build a new agent every week, against your own product work. Build it in one tool, and use the skills in any.

WEEK 1

Point AI at the right product problem

Start with the outcome you own, not the tool.

  • Pick a metric you own, and the product decision behind it
  • Load your product context: strategy, personas, funnel, tooling
  • Test the idea before you build the real thing
You’ll build

Your Product Brain, set up and in use

WEEK 2

Put an agent on a product workflow

Take one recurring piece of product work off your plate.

  • Connect to Linear, Notion and your analytics, and act on live data
  • Spec writing, feedback synthesis and the weekly metrics review
  • Read-only access by default, with a human gate before anything writes
You’ll build

Your first AI teammate, running a real task

WEEK 3

Run product discovery with an agent team

Own the whole outcome, from signal to decision.

  • Give each agent one role: usage, activation funnel, voice of customer
  • Hand off cleanly, so the evidence survives the chain
  • Add a lead agent that commits to a hypothesis and names the confounders
You’ll build

A multi-agent workflow running end to end

WEEK 4

Trust agents with the calls that matter

Make the output good enough to act on.

  • A scorecard, so agents grade their own work
  • A critic agent that reviews against that standard
  • An evaluation set that proves quality holds at scale
You’ll build

Agents you can trust to run without you watching

The 12 ways AI work stalls

The 12 ways AI work stalls
Pointing it at the right thing
  1. 1Using AI to go faster on the wrong work
  2. 2Confusing output with outcomes
  3. 3Building before validating, skipping discovery
  4. 4Chasing tools instead of problems
Building it well
  1. 5Vague instructions, then blaming the agent
  2. 6Keeping the agent’s brain in the chat, not in files
  3. 7Giving agents autonomy before they earn trust
  4. 8Building a team where one agent would do
Trustworthy, and alive
  1. 9Trusting confident answers with nothing behind them
  2. 10Shipping on ‘it worked once’
  3. 11Ignoring security, privacy and governance
  4. 12Set and forget, with no monitoring after launch

Every agent points at an outcome, and you stay on the calls.

You’ll recognise a few. In four weeks, you’ll learn all twelve.

Speed was never the problem. Direction is.

Every agent you build here points at a number you own.

89% of leaders say AI sped up their work. 6% can point to an organisation-wide return. Source: Atlassian, State of Teams 2026.

The weekly cadence

Four weeks. Live support, every single week.

Live Recorded
Monday

Open office hour

Live · 1 hr · optional

Any questions or advice. Bring your team. Some weeks we theme it on a topic.

Wednesday

Open office hour

Live · 1 hr · optional

More live help. Go deeper on one area with us before Friday.

During the week

1-1 session

Live · 30 min

Time with Raj on your own strategy or implementation.

Friday AM

Live group session

Live · 1 hr

Recap the week, Show & Tell, and Q&A before the next module drops.

Friday PM

Module drop

Recorded · 1 hr

New learning, a Product use case built live, plus homework to build real things.

Total time commitment: 3 hours per week + homework

More capacity, no new hires

Cover a bigger surface area without growing the team. The leverage comes from agents, not headcount.

Is this for you?

Built for product leaders who want their team building, not experimenting.

It works best for product people who can commit 3 hours a week plus homework, and would rather build than read.

This is for you if
  • You are a CPO, VP, Director or Lead setting the standard for your function
  • You want your whole product org operating with AI leverage, not a scattering of one-off experiments
  • You can carve out 3 hours a week, plus homework, for 4 weeks
  • You would rather your team owned this capability than a vendor
  • You learn by doing, and by showing your work to peers
  • You want evidence pulled and synthesised in minutes, not days
This is not for you if
  • You want someone else to build the agents for you (we do that too — see Consulting)
  • You are already shipping multiple AI agents per month
  • You cannot dedicate 3 hours a week for 4 weeks
  • You are looking for a strategy deck without hands-on work
  • You want recorded videos with no live access
Choose your plan

Three ways to join.

Best value

Team

£3,000

+ VAT

The Product leader signs up and brings up to 3 team members onto the course.

  • 4 seats total — you + 3
  • 4 live group sessions + 4 recorded modules
  • Weekly private team coaching
  • 2 weekly office hours + group Q&A
  • Coursework and community
Apply for the next cohort →
 

Individual with 1-1 coaching

£1,500

+ VAT

One seat, with dedicated coaching to build agents around your own workflows.

  • 1 seat
  • 4 live group sessions + 4 recorded modules
  • Weekly 30-min 1-1 coaching
  • 2 weekly office hours + group Q&A
  • Coursework and community
Apply for the next cohort →
 

Individual

£1,000

+ VAT

One seat on the course, learning and building alongside the group.

  • 1 seat
  • 4 live group sessions + 4 recorded modules
  • No private 1-1 coaching
  • 2 weekly office hours + group Q&A
  • Coursework and community
Apply for the next cohort →

All three plans run in the same cohort. Team pricing covers up to four seats — for a smaller group, ask us about individual pricing.

FAQ

What you’d want to know.

Where does our data go?
Agents connect with read-only access by default, so each one reaches only the data it needs and nothing more. Before you connect anything you map what flows where, and which model provider sees it. Every agent has a human gate: it presents its work and stops, so nothing writes, sends, files a ticket or contacts a customer without you approving it. You stay on the calls, by design, not as an afterthought.
Do we need to be technical?
No engineering team needed. If you can write a clear brief, you can build these agents, and we build alongside you every step of the way. Briefing an agent well is the real skill, and it is the one you practise every week.
How much time per week?
About 2 to 3 hours a week, plus homework: a 1-hour recorded module to watch in your own time, the live group session, and your 30-min 1-1. Optional office hours on Monday and Wednesday are there on top, if you want them.
Is this the right edition for me?
This one is built for product teams, around product work: discovery, experiments, activation and retention. If you lead marketing, take the Marketing edition. If you are a founder or CEO, take the Founders and CEOs edition. Same method, different worked examples, and you will get more out of the one aimed at your job.

More questions? Email Raj any time.

Next cohort

Join the next cohort.

Four weeks, live. Places per cohort are limited, so apply early to secure your team’s seats.