Forthcoming from Springer · 2027
How AI and Multi-Agent Systems Are Transforming Business
When AI begins to execute work, the company itself has to change.
A management framework for organizations moving from AI that generates answers to systems that pursue goals, use tools, decide and act inside defined boundaries. Written for executives, function heads and transformation leaders responsible for introducing those systems into real operations.
Explore the book on Springer Springer Cham · February 2027 · ISBN 978-3-032-46437-8
Agentic transformation is not a technology rollout. It is the deliberate redesign of work, authority, accountability and control.
The technology changes who can execute the work. It does not decide what the work is for, who answers for it, or what the organization is willing to carry if it goes wrong. Those decisions do not go away when the work moves into a system, and they do not make themselves.
The book follows one argument through four shifts, each of which changes something a manager is already responsible for.
From tools
Software moves beyond producing answers to completing defined units of work. As that work grows, specialized agents have to work together, which makes delegation and coordination management problems rather than purchases.
From human execution
When an agent holds the loop, the human job moves upward: setting goals, drawing boundaries, handling exceptions and owning outcomes that cross functions.
From pilots
Autonomy is granted on evidence rather than reached by maturity. It has to be bounded, recorded, independently checked, and withdrawable when the evidence changes.
From software products
Once software completes work rather than supporting it, a firm can reconsider what it sells, how that result is priced, and how value and risk are shared with the customer.
Alongside the argument, the book develops eleven management instruments for judging delegation, assigning ownership, governing autonomy and designing outcome-based offers. Among them:
The argument combines organizational research with operating experience. Nine first-person notes record what happened in AI-driven systems the author's team has built and run for financial institutions. A separate illustrative case, NovaCorp, is carried through every chapter and does not end in a clean success.
Autonomy is treated throughout as delegated authority granted on evidence, not as a ladder ending in maximum autonomy. For some decisions the correct permanent answer is supervised execution.
Alp Ustundag is a professor-practitioner and entrepreneur working at the intersection of artificial intelligence, financial analytics and organizational transformation. He is the founder and Managing Director of Navimod and leads Astera Finance by Navimod, its institutional platform for AI-driven investment strategies and decision systems.
He spent more than two decades at Istanbul Technical University as professor of industrial engineering and founding chair of its department of data engineering and business analytics, and is the author or editor of four Springer books.