Concept — illustrative data. A target-architecture concept. System names (M3, Vault, SAP) illustrate a typical integration pattern — not Øglænd's confirmed stack. Built to show where AI fits on a governed data foundation.
Øglænd Integration Hub
Your systems, one governed hub, your tools — a single bridge the AI tools build on.
One canonical model — every source maps to a single schema.
Your cloud.
Your data.
Your control.
The AI and the data platform run inside your own cloud tenant, on a managed enterprise platform. Nothing about how Øglænd works leaves your environment.
The AI and the data platform run in your environment — your data never leaves it.
Your records are never used to train models, and never pooled with anyone else.
Your own SSO / IAM controls access; the platform sits inside an isolated network.
Every connection and every store is encrypted — never plain-text, anywhere.
GDPR-aligned, with complete audit logging of every read and write.
Built on a hardened, enterprise-grade foundation your IT already trusts.
Enterprise-grade by design — built to satisfy a Hilti-scale security and compliance review.
Intelligence on the governed foundation
The same engine behind the Bid BOM Matcher, turned inward on your master data.
The Bid BOM Matcher's engine, applied to master data — one golden record per product.
LLM-assisted validation at ingest catches gaps and contradictions before they spread.
Natural-language across drawings, EPDs and certificates — find by meaning, not filename.
The matcher feeds clean, structured BOMs back into the hub as resolved records.
Golden-record health
Illustrative — completeness of the canonical product & material model.
What the foundation unlocks
Concept targets — directional, not measured.
All four app tabs — the Bid BOM Matcher, Pilot ROI, Market Projection and Market Opportunity — consume this one hub.
Data that flows through the hub
Modelled once, governed centrally, served to every app.
Products & SKUs
from M3The full catalogue — families, series, variants.
Bills of materials
from C4 + VaultAssemblies and component structures.
Stock & lead time
from M3Live availability and delivery windows.
Prices & discounts
from M3Customer pricing, agreements, currency.
CAD models
from Vault3D geometry and drawings per article.
Certifications & EPDs
from VaultStandards, approvals, environmental data.
Trusted by design, not by exception
The hub centralises data — so it also centralises control. One owner per entity, one gateway for access, encrypted in flight and at rest, with every read and write traced back to its source system.
Single source of truth
One owner per entity — the hub never invents data, it federates it.
- M3 owns commercials — price, stock, lead time, customer agreements.
- Vault owns engineering — CAD models, drawings, certifications, EPDs.
- C4 owns configuration — valid variants, rules, bill-of-materials logic.
Role-based access control
Who can read or write which entity, enforced at the policy gateway.
- Sales — read price, stock, catalogue; no engineering write.
- Engineering — read/write CAD, BOM structures and certifications.
- Partners & distributors — scoped read of public price + availability only.
Encryption everywhere
Data is protected on the wire and on disk — never plain-text.
- In transit — TLS 1.2+ on every source, hub and app connection.
- At rest — AES-256 on the hub store and its backups.
- Mutual auth (mTLS) on system-to-system source links.
Audit logging & lineage
Every read and write is traced back to its origin system.
- Immutable audit trail — who, what, when, from which system.
- Provenance — each field carries its source (M3 / Vault / C4).
- Tamper-evident logs streamed to a central SIEM.
Data quality & validation
The hub boundary is where messy source data becomes a golden record.
- Schema validation rejects malformed records at ingest.
- Dedup & matching collapses duplicates into one golden record.
- Reconciliation flags drift between source systems.
Compliance & privacy
Contacts and commercial data handled to GDPR and internal policy.
- GDPR — lawful basis, retention limits and erasure for contacts.
- Periodic access reviews — least-privilege re-certified each quarter.
- Secrets in a managed vault — rotated, never in code or config.
Stewardship & ownership
Accountability is assigned, not assumed — every domain has a named owner.
- Data steward per domain — A named owner for products, pricing, engineering and contacts — accountable for accuracy.
- Change-approval workflow — Master-data and price edits route through review before they reach the golden record.
- RACI on every entity — Responsible, accountable, consulted, informed — defined for each data domain.
Separation of duties
No single role can push a change end-to-end — checks are built in.
- Create, approve and release are distinct roles — no one person owns the full change.
- No self-approval of master-data or price changes; the maker can never be the checker.
- Release to partner-facing tiers requires a second authorised approver.
Retention schedule
Data is kept only as long as policy requires, then aged out.
- Operational data24 months
Live records and recent history kept hot for the apps.
- Audit & lineage7 years
Immutable trail retained for compliance and traceability.
Classification drives policy — Confidential is partner-invisible, encrypted and GDPR-governed; Public (catalogue, list price) is the only tier exposed to the distributor portal.
Today every tool re-integrates point-to-point with M3, Vault and C4. The hub is the single bridge: data is modelled, governed and served once. The Bid BOM Matcher already reads the product catalogue; once the hub is live it also reads real-time stock, lead time and price — turning a matched bid into a quotable, deliverable order.
- 1
Catalogue & BOM
DoneAlready done in the Bid BOM Matcher.
- 2
Hub: live stock + price
NextReal-time availability and pricing on every match.
- 3
Quote-to-order
One flow across all apps — matched bid to deliverable order.