ForgeKnowledge captures design decisions, engineering change rationale, fault diagnoses, and process parameters automatically from every Forge product. Full-text and semantic search. AS9100 Rev D clause 8.5.6 compliance as a direct output. Any artefact traces back to the decision that produced it.
Why ForgeKnowledge is different
"ForgeKnowledge captures decisions automatically from other Forge products — not from engineers filling in forms."
When ForgeOps creates a work order, ForgeKnowledge captures the decision: what triggered it, what data was used, what ForgeOps decided, and why. When ForgeCAD closes an engineering change, ForgeKnowledge captures the design rationale from the ECR fields. When ForgeMaint completes a PM, ForgeKnowledge captures the technician's findings. No engineer fills in a knowledge base entry.
"AS9100 Rev D clause 8.5.6 compliance is a direct output — not a documentation project."
AS9100 Rev D 8.5.6 requires organizations to maintain knowledge necessary for the operation of its processes and conformity of products. ForgeKnowledge provides the queryable record that clause 8.5.6 requires. When an auditor asks 'how do you manage organizational knowledge?', the answer is a URL to the ForgeKnowledge query interface and an exported decision log — not a procedure document.
"From any artefact, you can trace back to the decision that produced it — and every input that decision used."
Every artefact in Forge — a work order, a CAM program, a PM record, a compliance document — has a decision identifier. ForgeKnowledge stores the full context: what triggered the decision, who or what made it, what data was considered, what alternatives were rejected, and what the outcome was. Traceability is bidirectional.
ForgeKnowledge does not require engineers to maintain a knowledge base. It captures decisions as they are made, from the products that make them.
ForgeKnowledge subscribes to events from all Forge products via ForgeHub. Design decisions from ForgeCAD ECRs, fault diagnoses from ForgeOps, process parameters from ForgeMachine, audit findings from ForgeCompliance — all captured automatically.
Every captured decision is full-text indexed and semantically embedded. The machine, part number, subsystem, fault code, and date are extracted and stored as structured metadata alongside the full decision text.
Decisions are linked to the artefacts they produced — work orders, CAM programs, compliance records, PM schedules. Causal chains from ForgeHub are preserved: which event triggered which decision, which decision produced which artefact.
Engineers, quality managers, and auditors query ForgeKnowledge in plain English. 'Why was CONV-03 stopped last Tuesday?' 'What is the root cause history for Zone 4 drive faults?' Returns full decision context, not just a list of records.
Decision logs export as signed JSON, CSV, or PDF. ForgeCertify manifest entries are stored in ForgeKnowledge. External auditors receive an export package — they do not need vendor access.
Decisions are captured from ForgeCAD, ForgeOps, ForgeMaint, ForgeMachine, ForgeProcure, ForgeCompliance, and ForgeCertify — automatically, as they are made.
Full-text and semantic search across all captured decisions. 'Why was CONV-03 stopped last Tuesday?' returns the complete decision context — not a list of records to sift through.
From any artefact — work order, CAM program, compliance record — trace back to the decision that produced it, the data it used, and every downstream consequence.
Organizational knowledge management — as a direct output of ForgeKnowledge, not a documentation project. Auditors receive an export package, not a procedure document.
Automatic
Decision capture
No engineer fills in a knowledge base entry
AS9100 8.5.6
Compliance clause
Knowledge management — direct output
Bidirectional
Traceability
Any artefact → the decision that produced it
Full-text + semantic
Search
Plain English queries across all Forge products
ForgeKnowledge is not a knowledge base that engineers fill in. It is a decision archive that fills itself from the products engineers already use. Every fault diagnosis, design decision, and process parameter is captured, indexed, and queryable in plain English — the day it happens and five years later.