Ontology-driven Product Lifecycle Management control tower for Article's Product Development unit, tracking furniture products across a 52-week lifecycle from concept in Vietnam to landed in the US and Canada, with data-completeness scoring, AI task reminders, and automated risk detection.

ONTOLOGY / PRODUCT-DEVELOPMENT / VN→NA

Control Tower

SS26 line · 14 active SKUs wk 27 / 52

Portfolio Lifecycle — 52-week pipeline

concept → VN factory → landed US / CA
gate reviews auto-flagged
Completed Active stage At-risk Blocked Upcoming Current week

Risk radar

auto-detected across lifecycle
3 critical

Where risk concentrates

by stage
The semantic layer. Every fact about a product lives once, as a typed Object connected by Links. Departments read and write the same objects; Actions move a product through gates; AI Functions watch the graph for missing data & risk. This is what kills the VN→Canada inconsistency: there is no second copy of the truth to drift.

Object graph

core types & link types

Object type registry

Actions & AI Functions

writes back to the graph
Actions (guarded writes)
AI Functions (read the graph)

Object view

rid.product.sven-charme-tan
Stage 6 · Mass Production
Sven Sofa — Charme Tan
SKU
SVN-88-CHT
Category
Leather Sofa · 88"
Season
SS26 core
Factory
Bình Dương · Line 3
Markets
🇺🇸 US · 🇨🇦 CA
Target ship
wk 49 · Cát Lái → YVR

Data completeness

export-readiness index
4 fields block a clean US/CA customs + retail launch. Bilingual (EN/FR) content is the top gap for Canada.
18 complete 4 missing

    Linked objects

    one product, one graph — no drift
    AI Ops watches the graph so people don't have to remember. It knows which stage each product is in, which department owns the next step, and what data must exist before a gate. It nudges the right team, and catches bad data at entry — before it ships to Canada as an inconsistency.

    Next actions by department

    stage-aware reminders
    auto-assigned

    Data validation alerts

    caught at entry
    9 open

    Department workload — this week

    open tasks × SLA pressure