Before You Bet on AI, Get Flow Into Your Portfolio First
A 30-minute live experiment with the Xuviate Flow Simulator - and a surprise that should change how you think about portfolio improvement.
Everyone is rushing to layer AI and new tooling onto project delivery. But what if the biggest gains are already sitting underneath all of that - locked up in the policies that govern how work flows through the portfolio?
In just 30 minutes, Mark Geschke will run a full simulated year (260 working days) of a realistic, multi-team project portfolio - live, in the Xuviate Flow Simulator. We'll start from a baseline most leaders will instantly recognise: too much started, too little finished, deadlines slipping, and only around 10% of the potential financial value captured by year-end.
Then we change three things - one at a time:
- Limit work in progress at the portfolio level
- Adopt clear prioritisation rules across the organisation
- Enable targeted cross-team collaboration
No new hires. No new tools. No AI (yet). Still, by the end of the session you'll see throughput climb sharply, project durations shrink, every fixed-date deadline land on time, and total financial benefit more than double.
The results are deliberately uncomfortable. They suggest that for most organisations, the smartest next investment isn't another layer of sophistication - it's getting basic flow into the portfolio first. Until that's in place, even the best AI initiative is landing on a system that can't convert it into outcomes.
You'll leave with:
- A sharper intuition for the few decisions that genuinely move portfolio performance
- A clear before/after picture of what "fixing flow" actually looks like
- Free access to the Flow Simulator to run the same experiments - and your own - straight after the session
Duration: 30 minutes, live walkthrough.
Please address any questuions directly to Mathias Tölken (mathias@xuviate.com) or Martin Bertram (Martin.Bertram@pmi-gc.de)
Event Information
| Event Date | 27.05.2026 18:30 - 19:30 |
| Location | Online |


