This talk will describe how we can enable the businesses to manage software engineering efforts at scale in a financially sound way, all the while dealing with multiple projects/products or value streams, multiple events or deadlines, multiple stakeholders and multiple teams.

It will show how to resolve the conundrum of dependency management at scale by collapsing the combinatorial explosion of dependency networks into a simple linear queue.

It will describe how to instrument the organization in order to raise leading signals of oncoming trouble in order to trigger intense, frequent and engaging collaboration between all players (business and engineering), leading to an environment that unleashes the organization's collective intelligence via co-creation and social learning.

The talk will highlight techniques employed to foster intense collaboration between engineering and business, at scale.

It has been employed, among others, in a case in the automotive industry involving an organization with 8.000 people, half of which in 120 engineering teams distributed globally across the world, dealing with on average 400 projects in process at any time, and 70,000 change requests per month. The focused approached allowed the organization to improve it's operational throughput by 40% in a matter of months after inception.

The approach is data driven, and takes into account: Flow metrics, Buffer Burn Rate, Virtual Queue Sizes, Operational Throughput, Financial Throughput, Financial Throughput Rate.

The techniques used are a re-elaboration of the Theory of Constraints, taking elements of Drum-Buffer-Rope scheduling and Critical-Chain Project Management, and combining them with queuing theory and Little's Law.

The metrics and signals produced are also used to trigger collective psychological flow states leading to sustainable and optimal human experience at work.