EPU Product Intelligence

1. From product record to product intelligence

The strategic value of EPU is not the identifier itself. It is the ability to use that stable identity as the entry point into a governed product graph.

An EPU can therefore answer complex questions through its connections without storing the answers as mutable fields on the EPU identity record.

Question
   ↓
EPU
   ↓
authorized graph traversal
   ↓
identity + composition + supplier + production + event + evidence + calculation graph
   ↓
answer + provenance + confidence + effective context

Every material answer should be capable of returning not only a value but also the path and evidence supporting it.

2. Supply-chain intelligence

Examples include:

• Which manufacturers supply the parts in this product?
• Which Tier-1, Tier-2 and deeper-tier suppliers contribute to this EPU?
• Which supplier provides a specific component, material or substance?
• Which suppliers contribute more than a defined percentage of this product's GHG footprint?
• Which Tier-2 suppliers have unverified recycled-content claims?
• Which suppliers have missing or expired evidence?
• Which product batches used material from supplier X?
• Which products would be affected if supplier X became unavailable?
• Which components have single-source dependency risk?
• Which supplier contribution would most improve the confidence of the current footprint?

3. Composition and chemical intelligence

Examples include:

• What is this product made from down to material or substance level?
• Which BOM components contain substances of very high concern?
• Which substances lack sufficient declarations or evidence?
• Which supplier declarations support a REACH or SCIP-related claim?
• What percentage of the product is recycled, renewable, recyclable or hazardous material?
• Which material substitutions could reduce environmental impact without violating product constraints?

4. Carbon, PEF and lifecycle intelligence

Examples include:

• What is the current approved GHG footprint of this product?
• What is the footprint for a particular batch, production location or scenario?
• Which component dominates the product carbon footprint?
• Which lifecycle stage contributes most to climate impact?
• What changed between PCF v3 and PCF v4?
• Which emission factors, datasets and methods produced the result?
• Which MICE micro-engines participated in the calculation?
• Which data points are measured, supplier-specific, estimated or modeled?
• How would the result change if primary supplier data replaced secondary data?
• Which uncertainty contributors dominate the footprint confidence interval?
• How does the product perform under an alternative PEF/LCA scenario?

5. Logistics and freight intelligence

The EPU graph should support logistics as a first-class analytical context rather than merely a static footprint line item.

Examples include:

• Can we optimize the freight routes for the components in this product?
• Which inbound component routes contribute most to product transport emissions?
• Can a supplier, port, warehouse, mode or consolidation point be changed to reduce emissions?
• What is the GHG/cost/service trade-off between air, road, rail and sea scenarios?
• Could shipments from multiple suppliers be consolidated?
• Which route changes reduce emissions without violating lead-time, capacity or production constraints?
• Which component sourcing decisions create avoidable transport distance?
• What would the product footprint be if final assembly moved to another facility?
• Which freight lanes have the highest emissions intensity or uncertainty?

A freight optimization result is not an EPU attribute. It is a governed analytical result derived from product composition, supplier locations, logistics events, transport modes, distances, constraints, factors and relevant MICE optimization/calculation engines.

6. Circularity, repair and end-of-life intelligence

Examples include:

• Which parts are replaceable or repairable?
• Which spare parts are currently available?
• What is the product's repairability score and what evidence supports it?
• Which design changes would improve repairability?
• What percentage of the product can be technically recycled?
• Which materials prevent higher recyclability?
• Which recycler can process this product or its components in a specified geography?
• What end-of-life route produces the lowest expected environmental impact?
• Can components be reused, remanufactured or harvested before recycling?
• How does refurbishment affect the lifecycle footprint per functional unit?

7. Evidence and assurance intelligence

Examples include:

• Which evidence supports the 37% recycled-content claim?
• Who supplied that evidence?
• When was it issued and when does it expire?
• Has it been independently verified?
• Which verifier reviewed the underlying calculation or evidence?
• Which claims depend on estimated rather than primary data?
• Which product disclosures would become invalid if a certificate expired today?
• What is the TrustGate status of this product, component, supplier contribution or footprint result?
• Can the historical result be reproduced using the exact evidence and calculation versions available at that time?

8. Compliance and disclosure intelligence

Examples include:

• Is this product ready for a required disclosure profile?
• Which mandatory fields are missing?
• Which requirements are satisfied by authoritative source data?
• Which requirements depend on supplier contributions?
• Which disclosures are supported by verified evidence?
• Which regulatory requirement changed since the last disclosure version?
• What information can be shown publicly, to a verifier, to a regulator, to a recycler or only internally?

Pergamum Pulse can eventually provide authoritative regulatory requirement context to disclosure profiles while EPU remains regulation-neutral at the identity layer.

9. Design and optimization intelligence

Once composition, suppliers, logistics, processes, evidence and footprints are connected, EPU can support decision intelligence before a product is manufactured.

Examples include:

• Which component redesign gives the largest expected GHG reduction?
• What happens if virgin aluminium is replaced with higher-recycled-content aluminium?
• Which supplier substitution improves footprint without reducing evidence quality?
• Can the BOM be optimized for carbon, circularity, cost and compliance simultaneously?
• Which design change provides the best environmental improvement per unit cost?
• Where should production occur to minimize lifecycle impact subject to capacity and lead-time constraints?
• Which missing data should we collect first to reduce decision uncertainty most efficiently?

10. Acquisition and readiness intelligence

EPU should support questions about what information ought to be acquired before it becomes urgent.

Examples include:

• Which sustainability information should we request during this supplier onboarding?
• What should be requested with this RFQ or purchase order?
• Which likely future requirements are not yet satisfied for this product?
• Which missing information can already be resolved from supplier-connected infrastructure?
• Which supplier should be asked for each unresolved datum?
• Which data request has the highest expected value before contract award?
• Which evidence will expire before the next planned purchase or disclosure?
• Which new Pergamum requirement creates acquisition work for this product?
• How many supplier requests can be avoided through existing reusable product data?

These answers are governed requirement/gap/acquisition results, not mutable EPU identity attributes.

11. Procurement and negotiation intelligence

EPU should support procurement decisions using comparable sustainability results and governed financial-impact calculations.

Examples include:

• Which technically qualified supplier offer has the lowest carbon-adjusted acquisition cost?
• Are the candidate PCFs sufficiently comparable to rank?
• Which supplier has the lowest verified comparable PCF?
• At our internal carbon price, is a lower-carbon product premium justified?
• At what internal carbon price does Supplier B become economically preferable to Supplier A?
• What nominal discount would Supplier A need to compensate for its higher PCF?
• How would freight change the delivered sustainability-adjusted cost?
• Could uncertainty in either supplier PCF reverse the ranking?
• Which missing supplier evidence is most valuable to obtain before award?
• Which supplier substitution reduces Scope 3 emissions without exceeding the permitted adjusted-cost increase?

A procurement answer must preserve the distinction between contractual price, internal carbon/shadow cost and modeled adjusted decision metrics.

Product Sustainability Comparability must precede footprint ranking. Functional unit, product equivalence, system boundary, methodology, allocation, period, geography, data quality, uncertainty and TrustGate state are material comparison context.

12. Query response contract

Product intelligence should not return unexplained numbers. A mature response should be capable of exposing:

Answer
├── EPU context
├── scope / functional unit
├── effective date or production context
├── graph traversal path
├── source identities
├── evidence references
├── calculation/result versions
├── MICE engine lineage
├── method / factor / dataset versions
├── uncertainty / confidence
├── TrustGate status
├── comparability assessment where applicable
├── financial policy refs where applicable
├── access-policy filtering
└── generated-at timestamp

This allows the same EPU graph to serve interactive ZAYAZ interfaces, APIs, AI assistants, auditors, procurement decisions, DPP disclosures and optimization engines without sacrificing traceability.

13. Product intelligence principle

A product intelligence answer is a governed traversal and/or computation over the EPU Product Graph, not an unsupported AI-generated assertion.

AI can interpret questions, plan graph traversals, explain results and propose scenarios. Authoritative product facts, calculations, procurement comparisons and compliance conclusions must remain grounded in governed data, evidence, methods, policies and provenance.