Overview / Architecture
Architecture
A mental model for IT and operations, not a features list.
PD:OS is intentionally independent of product and project systems — designed to extend what you've already invested in, not replace it.
The layers
Four layers, one reference flow.
Product assets → evidence → learning → judgement → recommendations → accountable decisions → new evidence.
Authoritative facts, transactions, specifications and execution records remain where they belong.
The independent operating layer that stewards evidence, Product Learning Cycles, inheritance and governed recommendations.
AI accesses governed evidence to explain WHAT, SO WHAT and WHAT NEXT. It does not command.
People make accountable decisions and create the next cycle of evidence.
Data boundaries
Integration is a controlled relationship, not a central data grab.
BOUNDARY
Product IP
Design detail, specifications and proprietary product data stay governed by their own systems and permissions.
BOUNDARY
Confidential project evidence
Programme-specific evidence stays scoped to the ecosystem that created it.
BOUNDARY
Transferable judgement
Only patterns and judgement cleared for reuse compound across products and agents.
Honesty convention
Current implementation vs. future capability.
The platform evolves. This page stays honest about what's running today versus what's designed but not yet live.
Evidence capture with confirmation
Propose → confirm → log learning, tied to real evidence, retrievable across products and agents.
Explainable recommendations
WHAT? / SO WHAT? / WHAT NEXT? structure, with stated confidence and assumptions.
Broader connector ecosystem
Deeper native connectors for additional PLM, MES and QMS platforms beyond what's integrated today.
Cross-client governed collaboration
IP and access policy for transferable judgement across separate customer organisations — a named open decision, not yet resolved.