Vested or Bust — Renegotiating the Apex 3PL Deal
A four-round, advanced strategic-sourcing simulation set inside Cumberland Health Brands (CHB), a Nashville USD 820M consumer-health manufacturer…
Sobre la simulación
The learner is a key-account coordinator at a fictional Valencia packaging distributor who has tried AI agents for months without real value, and whose manager wants to know by January whether they save hours.
Decision 1, Monday 4 October 2027: which three flows to automate (the three from a vendor demo, a mix, or the three a time log points to), how to measure (by feel, the vendor dashboards, or a baseline log) and which tool (five apps on free trials, a personal premium plan, or the corporate assistant with its agent builder and data-processing contract).
Decision 2: how much autonomy (full, drafts only, or by the cost of an error), which instructions (one line, a marketplace template, or a one-page playbook with examples, rules and stop conditions) and which data each agent sees (everything, least privilege, or a rule forwarding all work email to a free app on a personal account).
Decision 3, after an agent sends last year tariff to a client and an email with hidden instructions tries to extract the price list: what to do about the error (nothing, switch everything off, or a review), how to supervise (a glance, read everything, or by risk) and what to do with what does not work (add agents, keep them, or kill the weakest using the log).
Decision 4: roll out to the team, stop, or pilot with one colleague; what to tell the manager (nothing, a felt ten hours, or the log); and how to keep it going (nothing, more tools and a course, or a weekly review).
A deterministic model turns these into gross savings, supervision, upkeep, rework, errors reaching clients and the share of the stack still in use after sixteen weeks, and drives net hours gained, errors, agents in use, the gap between believed and real savings, and own spending. The score combines the quality of each decision with those outcomes and is capped when work email is forwarded to a personal app.
Spanish (primary, Spain) and English.
Para quién es
Una simulación de nivel intermedio para participantes que ya conocen los conceptos básicos de la materia, pensada para programas de máster y MBA.
Cómo funciona una sesión
- El docente crea una sesión desde el panel de Eureka e invita a los participantes.
- Los participantes juegan 4 rondas. En cada una envían sus decisiones y la simulación calcula los resultados.
- La simulación se puede jugar en: English, Español.
- El docente sigue el progreso y los resultados de cada participante desde el panel, y usa los resultados de la clase para el debriefing.