Case Study · FedEx · 2024–Present

AI & agentic workflows inside a $75B activity-based costing model

RoleCo-lead, enterprise ABC system

TeamCosting SMEs + data scientists

Scale~$75B cost base

The problem

[Draft placeholder] Allocating an enterprise cost base of roughly $75B requires cost logic that is accurate, explainable, and fast to iterate. Legacy approaches to forming and validating that logic were manual and slow relative to the pace the business needed.

What I built

[Draft] Co-leading development of a new enterprise-wide activity-based costing system.
[Draft] Deploying AI directly into the cost model — accelerating how cost logics are formed, experimented on, and validated.
[Draft] Building agentic-AI workflow prototypes and validation scripts for the costing process.
[Draft] Managing a team of costing subject-matter experts and data scientists running the experimentation loop.

What changed

[Draft placeholder — quantified impact coming: time saved per model cycle, validation speed, dollars surfaced.]

Why it matters

[Draft placeholder] This is what applied AI in supply chain finance actually looks like: not a demo, but production cost logic at one of the largest operational scales in the world.