Judgment & relationships
Work that needs human accountability and the ability to navigate ambiguity no prompt can fully specify. Owner: human, always.
FIG 01Quarterly executive cohort · Next: 18 September 2026 · Singapore
Your org wasn’t built for agents. The hybrid organisation is.
You have a mandate. You’ve spent the budget. The org hasn’t moved. MIT’s 2025 study found the barrier isn’t the technology. It’s organisational design. We run 40% of our workforce as agents in production. In one day you’ll walk that live fleet, spec your first agent, and leave with the numbers to defend it upstairs.
Date18 September 2026
LocationSingapore · in-person
SeatsStrictly limited
InvestmentS$1,990
FIG 02The numbers
Real economics, not a forecast.
40%
of MING's workforce runs as agents in production today.
95%
of GenAI pilots show zero measurable P&L impact.
Source · MIT NANDA, State of AI in Business 2025
67% vs 33%
deployment success with an external partner vs building alone.
Source · MIT NANDA, State of AI in Business 2025
FIG 03The method
Every role breaks into
three kinds of work.
A
B
Structured expert work
Reporting, analysis, meeting prep, documentation. Skilled work that follows patterns. Owner: designed. Human or agent, by role.
C
Routine operations
Data prep, inbox triage, status tracking, about 40% of an expert's week. Owner: agent. This is where you start.
The rule when in doubt: two narrow agents beat one fat one, in output quality and in maintenance cost.
40%-of-week figure: MING role-decomposition data, enterprise engagements 2024–26
FIG 04What you leave with
Four things, in your hands
by the end of the day.
A
Your first agent, spec'd and costed
Built in the afternoon lab against a real task from your org. A written spec: role, inputs, outputs, autonomy level, and the human baseline it replaces, sized on time, quality and cost.
B
The numbers that make the case
Real per-task numbers from MING's own fleet: compute, build and review. Plus third-party proof for any sceptic.
C
A reusable method
The JD-to-three-buckets tool: any role broken into what stays human, what gets AI-assisted, and what an agent owns. Walk out knowing how to apply it to your next 10 tasks.
D
Cover to defend it internally
The governance answers that get IT, security and risk to yes. Answers to the five objections you'll face on Monday. A first move sized small enough to authorise without a formal business case.
FIG 05The day
From why it didn't move
to your 90-day path.
Morning
Why the org didn't move, and what real looks like
The mechanism: why individual speed never became org throughput. Then a live walk of MING's fleet: real agents, real tasks, real per-task economics on one org chart.
Afternoon
Build lab: spec your first agent
Small groups of senior peers. You pick the task, apply the method, write the spec. You leave with a real first move, not a homework slide.
Close
Your 90-day path from here
Champion plus one revenue-adjacent use case. Sandbox to pilot to scale. How to bring IT and risk with you, and defend the project after the room empties.
+2 wks
Online checkpoint
The cohort reconvenes to show what they shipped. The peer group sticks. It leads into an optional 1:1 review of your first build.
FIG 06Who’s leading it
Led by an operator,
not a theorist.
SMPhoto · Sebastian Mueller
Sebastian Mueller
Founding Partner · MING Labs
Fifteen years inside enterprise systems: 62 Fortune 500 companies and industrial leaders like Bosch, Voith, MANN+HUMMEL and Zeiss. 250+ projects, 35+ corporate ventures built. Co-founded MING Labs in 2011; bootstrapped, profitable, four offices.
He doesn’t teach the hybrid organisation. He operates one. 40% of MING’s workforce runs as agents on one org chart, plus an agent-only venture. On 18 September he walks you through the live fleet: real tasks, real governance, real per-task economics, not slides.
“Everyone leaves with a first agent spec’d and costed, and a 90-day path to get it running.”
62 Fortune 500 clients
35+ corporate ventures
Articles read 4M+ times
40% agent workforce
LinkedIn ↗
FIG 08Is this you?
A strictly limited room.
The right people in it.
Right seat
A mandate to deliver
Head of Innovation, Digital or Transformation. Senior Director with a personal AI mandate in a large enterprise. Sectors: industrials, energy, financial services, mobility.
Right moment
Spend, but no movement
You have budget and a mandate. You've rolled out tools. The org hasn't moved. You need the first real win, with proof leadership will accept.
Not this room
A different conversation
CxO / board: that's a separate briefing. Engineers and practitioners: that's the POC. Anyone without an active mandate to deliver measurable results.
One cohort, no competitors. We screen every sign-up before confirming a seat: no two competing companies share a session. It is why confirmation takes up to 48 hours instead of being instant.
FIG 09Secure your seat
18 September. Singapore.
Strictly limited.
S$1,990
Per seat
Book directly
Leave with your first agent: spec’d, costed, ready to build.
Secure your seat →No payment now: sign up, we screen within 48 hours to keep the room non-competing, and accepted participants receive a payment link (card or invoice).
If you have any questions, want to place a team order, or need to book for more than one person, please contact Sebastian: sebastian.mueller@minglabs.com.
The fine print. Seat transferable to a colleague anytime · full refund up to 14 days before · venue confirmed at least 3 weeks out · invoice available on request.
We run this quarterly. Small cohorts, non-competing enterprises, different cities. Each session is independent; 18 September is the next date. If it’s full:
Join the Q4 waitlist →
Join the Q4 waitlist →