A lab where senior operators redesign the workflows AI is changing. You already use AI individually. Your team does too. But the work itself hasn’t changed yet: reporting is still manual, coordination is still painful, handoffs are still messy. That gap is what we work on. Minimal theory. Maximum implementation.
Request an invitationTopics span two sessions: built in one, tested inside your company in between, presented back in the next. Four full build cycles across five weeks.
The processes internal to your organisation: 1:1s, performance conversations, working with peers, managing up. You pick the angle that fits your role.
How people in your org know what’s happening: cross-functional coordination, post-meeting updates, decision flow. Qualitative visibility, not metrics.
Documentation as AI infrastructure: where information sits, how it’s structured, and whether your AI can reach it. This is where you stop knowledge living in people’s heads and set it up so AI can build on it.
Participants propose the final topic, the room picks together, and you self-allocate into build groups around the problems that matter most.
Concept material lives in async pre-work. Live time is for building. Each session opens with presentations from the previous build cycle.
A short framing of the topic and the AI pattern underneath it, then one case study discussed together to create common ground.
Breakouts in small teams: who’s involved, how it works today, where it breaks, the outcome you want. Shared in 60-second rounds.
Building starts live, in your real environment, with whatever tool fits: an agent, a scheduled task, an agentic workflow.
Run the redesign inside your company. An AI interviews you about the build and writes it up, so every redesign is documented consistently.
The full time cost, stated plainly: two hours live each week for five weeks, plus around one hour of pre-work per session. Roughly fifteen hours, total. The format exists so you don’t spend the rest of your week trying to keep up.
Every operator keeps a participant page the whole cohort can read: the workflows you’re redesigning, the experiments, what worked, what broke. We build with each other, and everyone in the room has been handpicked or recommended, so everyone has something to teach.
The documented builds feed a cohort AI you can query for patterns: how someone automated 1:1 follow-ups, what worked for managing up, which visibility redesigns held. At the end of the programme, you post what you built and what you learned, on LinkedIn, internal Slack, or your all-hands. Accountability through visibility is part of how this works.
Your seat includes 12 months of Workbench membership, starting the day your cohort does.
A private community of operators rebuilding how companies work for the AI era. Ask, compare, share what you’re shipping.
A live two-hour session each quarter, deep on one area: finance, marketing, operations.
Bring two colleagues, free, to every deep dive. Your team sees the method without buying a seat.
The Workbench is about redesigning workflows. If that’s within your reach, even just for your own team, these five weeks will be worth your time.
You don’t need to be an expert, but you should be working with today’s tools, agents included. The cohort builds together instead of covering basics.
Sound like you? Send the request. We’ll have a short conversation, and you’ll get a straight answer.