About

How I work

I’m Pedro Alizo. For customer-facing engineering, I use a fixed sequence: discover the user, process, data, and control boundary; make the checkable path visible; prototype narrowly with synthetic or permitted data; evaluate before anyone proposes a model-assisted narrative layer; then hand off with owners and review states clear. Discovery without controls is theater. Prototypes without evaluation are demos.

Finance and systems change

My background runs from planning and reporting to leading regional teams and implementing SAP across Latin America. Working across those layers made one thing clear: finance systems earn trust when definitions, ownership, and the operating process fit together. That domain fluency is what I bring into forward-deployed and applied-AI engineering work—not a claim that I already hold those titles.

Evidence and writing

The statement workbench focuses on visible calculations and missing-data states; the variance workbench models exception triage and reviewer decisions. Both use invented scenarios and demonstrate narrow behavior, not integrated production systems or measured outcomes. My writing examines finance AI through controls, economics, and organizational judgment; Start Here groups it by question. Engineering notes on the static workbench and the drift eval harness sit alongside the practitioner essays.

I’m based in Madrid and speak English and Spanish natively. This is a personal site and reflects my own work and views. For my work history, see the CV; for role conversations, use the contact page.