AI Agent Builder Workshop

Your team builds its own AI agents. On your real work, in four sessions.

Most AI training shows your team what the tools can do, and two weeks later nothing about the work has changed. The sprint is built the other way around.

Every attendee builds a working agent in the first session, maps their whole role in the second, aims agents at the goals that matter in the third, and prices the result in the fourth. You leave with working agents, not notes.

Book a call

What happens each session

One build per session. Homework between them.

Sessions run about 90 minutes, spread across two weeks so the team can run what they build between sessions. That gap is deliberate. An agent that survives a week of real use is an agent your team will keep.

Build your first agent. Four steps: create a folder, install the Agent Builder, describe one piece of repeatable work, and watch it become a working agent. No coding knowledge needed at any point. Walk-out: every attendee has built and run one working agent

Which agents should we build? Map each role into six areas of responsibility and six deliverables per area. The map becomes a backlog, so the team builds what matters next instead of whatever is in front of them. Walk-out: a 36-deliverable role map and a build backlog

Crush your goals. Shift from clearing busywork to going after what you actually want, including projects you had written off as unaffordable. Goals become projects, and projects become chains of agents working in sequence. Walk-out: ten goal projects ranked, with the first agent chain built

Execution ROI. Run the numbers. What the agents produce in a year, priced against what you would otherwise pay in hours or outside help, and a ranked plan for what to build after the sprint ends. Walk-out: an annual value figure and a next-quarter build plan

Two ways in

Standard, or built around your operations.

Same four sessions either way. The difference is what the exercises are made of.

Standard Sprint

$900 USD

first seat · $450 each additional

Proven exercises, your team's own first agents.

Four live sessions across two weeks
Every attendee builds a working agent in session one
Role maps and build backlogs for each attendee
The Agent Builder installed and yours to keep using
All session materials and recordings

Book the standard sprint

Customised Sprint

From $1,350 USD

first seat · $600 each additional

Your operations become the exercises. Final price confirmed with your outline, based on scope.

Everything in the standard sprint
Exercises built from your named operations, systems and documents
Your highest-value workflow becomes the session-one worked example
Operational agent chains designed and built live in session three
ROI session priced in your numbers, not illustrative ones
A customised outline before you commit

Request your custom outline

Three seats comes to $1,800 on the standard sprint, or from $2,550 customised. Four seats or more, ask about the team rate, though most teams that size are better served by the customised sprint. Payment is by invoice before the first session.

The practicals

What the sprint needs from you.

Format. Four live working sessions of about 90 minutes, run online over two weeks. Every session has hands-on building in it.

Time zones. B2M runs from New Zealand, so US and European working hours are both covered. Sessions are scheduled around your team's calendar, not ours.

Currency. Prices are in US dollars. Pay online in your preferred currency. Teams outside New Zealand pay no GST. New Zealand businesses are invoiced in New Zealand dollars with GST.

Software. Each attendee needs a Claude subscription and the Claude desktop app, from $20 a month. That subscription is the only cost outside the sprint fee.

Experience. Nobody writes code. If your team can write a work instruction for a new hire, they can build an agent.

Who runs it

Delivered by someone who builds these agents for clients every week.

The sprint is run by Simon Fawkes of B2M. The method is not theory. The same agent patterns taught in the sprint run live client work, from proposal generation for a marketing consultancy to documentation builds for a software team.

Simon works alone, and that is deliberate. The person who builds alongside your team is the person you talked to.

More about B2M

Before you ask

The questions that come up first.

What happens if an attendee misses a session? Every session is recorded and the materials are yours to keep. The build exercises work solo, so a missed session gets caught up before the next one.

Is this for individuals or teams? Both work. Teams tend to get more out of it, since the role maps from session two start to overlap and agents get shared between people.

Do the agents keep working afterwards? They run in your own Claude account, in your own folders, on your own files. Nothing sits on a B2M platform, and there is no ongoing fee to keep using what your team built.

How is this different from an AI training course? A course would leave your team with an understanding of agents. The sprint leaves them with a folder of agents that ran on this week's work.

Book a call