D-POAF® Operating Guide

AI is already part of software engineering. It helps teams frame requirements, generate code, review changes, and operate systems. The benefits are real. So is the governance problem that comes with them: when humans and AI shape the same delivery process, decisions can become difficult to reconstruct, responsibilities harder to trace, and evidence fragmented across tools.

D-POAF® "the Decentralized Proof-Oriented AI Framework" was developed to address that gap.

This Fourth Edition of the D-POAF Operating Guide presents a practical way to bring proof, traceability, decision authority, and living governance into AI-native software engineering. It introduces the Wave as the unit of governed progress, the three proof families : "Delivery, Value and Reliability" and the roles, rules, decision models, and registers that support them.

Through worked examples and a complete reference implementation, the book shows how D-POAF can coexist with Agile, DevOps, and MLOps, from a single team to organization-scale AI governance.

The objective is not to add another layer of process. It is to make the reasoning, evidence, and accountability behind software delivery easier to see, verify, and govern.