| 8eabeab8103aad3a… | Context relationships link a CXU to the entity or entities it concerns so that each knowledge unit is attached to the real-world subject it describes. | active | contextual | yes | |
| 2dcbedda61c17959… | The graph distinguishes three classes of relationships at a high level: context relationships, knowledge relationships, and scope relationships, so that graph structure captures different kinds of meaning and control. | active | contextual | yes | |
| 122fa99ad36eade4… | A CXU may belong to multiple sets, including universal sets or sets representing a regulatory regime, business unit, customer environment, region, or internal operating boundary, so that the same knowledge unit can participate in multiple scoped contexts. | active | contextual | yes | |
| e5d0360923f13f5c… | A naive RAG system might retrieve paragraphs mentioning 'batch record,' 'change,' or 'Packaging Line 3' and leave the user with text matches rather than governed graph-based resolution. | active | contextual | yes | |
| d320a4cad75d7579… | The retrieval walkthrough uses the question 'What governs changes to electronic batch records on Packaging Line 3?' as the example query for illustrating system behavior. | active | contextual | yes | |
| 9259a5e0c56817c0… | Results from graph and semantic retrieval are fused using weighted ranking so that structural relevance and semantic relevance reinforce each other rather than replacing one another. | active | contextual | yes | |
| 2e9dc37c05c96f20… | Semantic vector retrieval runs in parallel with whichever graph strategy is selected so that semantic evidence is gathered alongside structural retrieval. | active | contextual | yes | |
| f393c975804f3e20… | Free-form graph search dynamically generates graph search behavior for exploratory or less structured requests so that the architecture can reach beyond the template library while still using the governed graph. | active | contextual | yes | |
| 8791d29fa4b7bdb0… | Scoped graph traversal is useful when an answer must be limited to a specific policy regime, department, region, or tenant context so that results remain within the required scope. | active | contextual | yes | |
| 5dc2699f4b87ad02… | Scoped graph traversal constructs a scoped graph query when a request includes set, domain, or relationship constraints so that retrieval respects those boundaries. | active | contextual | yes | |
| 5cb091135812aed5… | Template graph retrieval uses pre-defined graph templates for common enterprise question patterns so that known intents such as governance, dependency, lineage, and requirements lookups are handled deterministically with bounded traversal. | active | contextual | yes | |
| 6dd68be7ef5cc050… | Pyrana uses three complementary retrieval strategies and fuses their outputs with semantic retrieval so that retrieval is not limited to a single search method. | active | contextual | yes | |
| f9e799e7bcf10d16… | Retrieval is described as the system's most differentiated subsystem because Pyrana does not treat every question as a similarity search over text and instead combines multiple retrieval strategies with semantic retrieval. | active | contextual | yes | |
| 954aef367e507b6c… | The graph is not merely a decoration around vector search; it is the structural model that makes scoped and explainable knowledge retrieval possible. | active | contextual | yes | |
| 1b4850c5192c713a… | In the example graph, governance determines which connected knowledge item should dominate when a regulated requirement and a local operational note both apply to the same entity. | active | contextual | yes | |
| 5a3eadc884f0bd18… | Even at small scale, the graph differs visibly from chunk retrieval because it can distinguish a regulated requirement from a local operational note and connect both to the same entity under governance. | active | contextual | yes | |
| 828bb510404efa83… | The stylized subgraph includes a real-world entity, a regulated rank-0 claim, a standard CXU, and a local note marked low priority so that different knowledge object types are represented in one example. | active | contextual | yes | |
| 187611953120787c… | The example subgraph is a stylized illustration of how a small corpus might look for a pharma operations question so that readers can visualize the graph structure in a domain-specific scenario. | active | contextual | yes | |
| 7da224845db758a4… | The resulting corpus can be traversed by policy, entity, lineage, and semantic similarity rather than by similarity alone so that retrieval uses structural and semantic paths together. | active | contextual | yes | |
| ab0e53b8cd6606d3… | Knowledge relationships include dependency, contradiction, specification, and governance links so that the graph can represent substantive interactions among knowledge units. | active | contextual | yes | |
| 12f62cc45454a787… | Context relationships link a CXU to the entity or entities it concerns so that knowledge units are attached to the real-world subjects they describe. | active | contextual | yes | |
| a805eccd05563068… | The graph distinguishes three high-level classes of relationships: context relationships, knowledge relationships, and scope relationships, so that different kinds of connections are modeled explicitly. | active | contextual | yes | |
| dfb863d3fb22a95a… | Scoped sets provide a practical path to multi-tenant knowledge management by allowing shared foundational knowledge to coexist with tenant- or department-specific corpora while preserving scoped retrieval and governance. | active | contextual | yes | |
| 7b1ef9bc46018f5b… | Sets create scoped subgraphs that let the system answer the same question differently when different domains, tenants, or policy frames are active so that retrieval can adapt to the active context. | active | contextual | yes | |
| 7274288294f32e10… | A CXU may belong to multiple sets, including universal sets or sets representing regulatory regimes, business units, customer environments, regions, or internal operating boundaries, so that the same knowledge unit can participate in multiple scoped contexts. | active | contextual | yes | |
| 9070817cc3c5f086… | Scope relationships express membership in sets, domains, or operating boundaries so that the system can determine the contextual boundaries within which a claim applies. | active | contextual | yes | |
| a0d2a404628e0b31… | Knowledge relationships include dependency, contradiction, specification, or governance so that claims can be connected according to their logical and control interactions. | active | contextual | yes | |
| c176c660affae982… | Context relationships link a CXU to the entity or entities it concerns so that the system knows what real-world objects the claim is about. | active | contextual | yes | |
| acedb9d14d443a0a… | At a high level, the graph distinguishes among three classes of relationships so that relationship semantics are organized into a defined taxonomy. | active | contextual | yes | |
| b64d058ef055ee18… | Shared foundational knowledge can coexist with tenant- or department-specific corpora while preserving scoped retrieval and governance so that common and local knowledge remain separable but usable together. | active | contextual | yes | |
| 65f7f98972fef8a6… | Scoped subgraphs create a practical path to multi-tenant knowledge management so that shared and tenant-specific knowledge can coexist within one system. | active | contextual | yes | |
| 0e440a19582f2068… | Sets create scoped subgraphs that let the system answer the same question differently when a different domain, tenant, or policy frame is active so that responses remain context-specific. | active | contextual | yes | |
| 9b408cc7aeb6e3d4… | Sets may be universal or may represent a regulatory regime, business unit, customer environment, region, or internal operating boundary so that scope can be modeled across many organizational dimensions. | active | contextual | yes | |
| a16258809c4a0947… | A CXU may belong to multiple sets so that the same knowledge unit can participate in more than one scope or operating context. | active | contextual | yes | |
| 131c46df0f845ac1… | Graph attachment makes relational enterprise questions first-class so that these questions are directly supported by the knowledge architecture rather than treated as indirect search tasks. | active | contextual | yes | |
| 483a2213f404a538… | Users may ask what evidence supports the current instruction for a workflow so that the system must connect instructions to their supporting evidence. | active | contextual | yes | |
| 11f2b5cc03a92d83… | Users may ask which requirements apply to one department but not another so that the system must represent differential applicability across organizational scopes. | active | contextual | yes | |
| 67370ed52026c6d8… | Users may ask what changed for a process after a policy update so that the system must track process knowledge over time relative to policy changes. | active | contextual | yes | |
| 0e5078cab70de841… | Users may ask which procedures govern a piece of equipment so that the system must support relationship-based retrieval between equipment and governing procedures. | active | contextual | yes | |
| e0255634d0b3d7fb… | Enterprise questions are usually relational rather than purely similarity-based so that users often need answers about relationships among entities, policies, and evidence. | active | contextual | yes | |
| cb6537422eb7ccec… | The graph model gives the system a stable representation of real-world organizational objects such as products, facilities, SOPs, regulations, systems, departments, controls, vendors, and process steps so that knowledge can be organized around persistent entities. | active | contextual | yes | |
| 1d7f36d1bc08fe8c… | CXUs do not live in isolation and instead attach to canonical entities and scoped knowledge sets so that the system knows where, how, and for whom a claim applies. | active | contextual | yes | |
| 60986ce1849a77c3… | The graph is the second foundation of the system so that the architecture relies on both CXUs and graph structure as core components. | active | contextual | yes | |
| a0961e51ed758434… | Without atomic claims, provenance, policy, contradiction handling, and learning all become much harder to govern so that atomicity is necessary for manageable knowledge governance. | active | contextual | yes | |
| d97d0f8296f530dd… | The CXU is the reason the rest of the architecture works so that the system’s broader capabilities depend on atomic knowledge units. | active | contextual | yes | |
| 43f677f615ce5517… | In Pyrana, governance begins at the unit of knowledge itself so that control is embedded directly in each knowledge unit rather than added later. | active | contextual | yes | |
| 89ed5ecb06357ed9… | Conventional retrieval pipelines usually apply governance after retrieval as an application-layer patch so that governance is external to the retrieved knowledge units. | active | contextual | yes | |
| fa9488c951811d38… | Regulated claims can require approval for lifecycle changes so that changes to regulated knowledge pass through explicit governance controls. | active | contextual | yes | |
| 169c44c64a3a8c66… | Foundational or policy-critical claims can be protected from autonomous modification so that especially sensitive knowledge is not changed automatically. | active | contextual | yes | |
| ef32b89bbd1f5806… | The classification model determines what the system may do with a claim so that system behavior is controlled by the claim’s classification status. | active | contextual | yes | |