Skip to content

Glossary

Canonical glossary for jinflow. Generated from terminology/terminology.yaml.

DE: Analytische Pyramide | FR: Pyramide analytique

The five-layer analytical stack from raw detection to actionable insight: Gold → Signal → Perspective → Thesis → Verdict.

Is: the conceptual stack that organizes all analytical output

Is not: a physical pipeline — each layer is compiled independently from YAML to SQL

Aliases: pyramid, pyramide, stack, analytical stack

DE: Kalibrierung

The process of measuring how accurately signals detect known defects, using recall, precision, and F1 against a defect manifest.

Aliases: calibration, kalibrierung, recall, precision, f1

DE: Vertrag | FR: Contrat

A JSON schema that defines the required fields, types, and constraints for a layer’s output, enforcing interface stability.

Aliases: contract, vertrag, contrat, schema

A human-curated narrative grouping of related Subject Matter — the story layer on top of atomic expert knowledge.

Is: a narrative wrapper that connects related Subject Matter

Is not: a perspective or automated grouping. Dossiers are hand-curated by a named curator

Aliases: dossier, bitbundle, bundle, use case

DE: Entität | FR: Entité

A domain object with identity — the thing being examined. Core entities: Case, Procedure, Material, Cost Center, Movement, Usage, Billing Event.

Aliases: entity, entität, entité, dimension

DE: Evidenz | FR: Évidence

A weighted reference to a signal or Subject Matter entry that contributes to a thesis evaluation. Each evidence link has a role (primary, supporting, context, counter).

Is: a scored link between a thesis and its supporting/contradicting data

Is not: the raw data itself. Evidence is the connection, not the finding

Aliases: evidence, evidenz, preuve, evidence chain

DE: Evidenz-Score | FR: Score d’évidence

A numeric value (0.0–1.0) that summarizes the total weighted evidence for a thesis, determining its thesis status.

Aliases: evidence score, evidenz-score

DE: Befund | FR: Constat

A single anomaly or pattern instance detected by a signal — with a continuous score (0-100), affected entity, and impact assessment.

Is: a specific, attributable detection event

Is not: a “result” or “alert.” Use “finding” consistently — never “signal result”

Aliases: finding, signal instance, befund, résultat, constat

DE: Gesundheitsscore | FR: Score de santé

An entity-level numeric score (0–100) produced by perspectives, aggregating scores and finding count across multiple signals.

Aliases: health score, gesundheitsscore

DE: Auswirkung

The estimated exposure associated with a finding, expressed as a value with unit and dimension. Generalizes the former money_at_risk field to cover financial, operational, quality, cultural, and strategic effects.

Aliases: impact, money at risk, risikobetrag, montant à risque, money_at_risk, mar, auswirkung, exposure

Generic term for any declarative analytical artifact: signal, perspective, thesis, verdict, Subject Matter, or Dossier.

Is: the umbrella term for the six artifact types

Is not: used in user-facing UI. The Explorer shows signals, theses, etc. by their specific names

Aliases: instrument, artifact, artefakt

FR: Interprétation

A human-readable explanation of a finding — what happened, why it matters, and what to look at next.

Aliases: interpretation, interprétation, erklärung

DE: Perspektive

An entity-level health score aggregating findings from multiple signals. Perspectives answer: “How healthy is this entity overall?”

Is: a structured, multi-signal aggregation with scoring rules

Is not: a subjective interpretation. Perspectives are deterministic given their source signals

Aliases: perspective, assessment, bewertung, évaluation

DE: Präsentation | FR: Présentation

A fullscreen, executive-ready view of a thesis evaluation with narrative, trend charts, and evidence chain visualization.

Aliases: presentation, präsentation, executive brief

DE: Risikomenge | FR: Quantité à risque

The estimated quantity exposure of a finding — how many units are affected. Complements impact for volume-sensitive signals.

Aliases: qty at risk, risikomenge, quantité à risque, quantity

FR: Registre

A dbt model (SQL table) storing metadata for all instruments of a given type, generated by the compiler from YAML definitions.

Aliases: registry, registre, register

DE: Bericht | FR: Rapport

A structured document generated from pipeline data. Types: data quality, financial risk, analytics readiness, and health.

Aliases: report, bericht, rapport, pdf

DE: Risikostufe | FR: Niveau de risque

A tiered financial risk classification assigned to each signal, used to prioritize findings in reports and the Explorer.

Aliases: risk tier, risikostufe, niveau de risque

DE: Grundursache | FR: Cause racine

The underlying reason why a problem exists. Structured into six categories: process, system, data quality, behavioral, structural, external.

Aliases: root cause, grundursache, cause racine, ursache

DE: Punktzahl

A continuous numeric value (0-100) quantifying the significance of a signal finding. Replaces the former discrete severity levels (high/medium/low). Display labels are derived from score ranges for human readability.

Aliases: score, severity, schweregrad, sévérité, punktzahl, significance score

A declarative diagnostic query that detects a specific pattern in Gold data. Signals produce standardized findings with a continuous score (0-100), impact, polarity, and direction.

Is: a repeatable, question-driven analytical intervention with rich output metadata

Is not: an ad-hoc query or a report. Signals are defined declaratively, compiled to SQL, and produce contract-compliant output

Aliases: signal, probe, sonde, diagnostik

DE: Studio / Salon / Labor

Three activity rooms with distinct tempos: Studio (build, days-to-weeks, async), Salon (meet, minutes-to-hours, sync), Lab (test, hours-to-days, semi-sync). The Explorer names its spaces with this vocabulary — Price Lab, Map Salon, Spend Studio.

Aliases: studio, salon, lab, three rooms

An atomic, attributed piece of subject matter expert knowledge — a single statement, insight, or known exception contributed by a named expert.

Is: structured, first-class expert knowledge with optional data validation

Is not: a comment, annotation, or wiki page. Subject Matter is versioned artifact with identity, lifecycle, and attribution

Aliases: subject matter, smebit, statement, check, expert knowledge, expertenwissen, connaissance

DE: Lieferorganisation | FR: Organisation de fourniture

The smallest organizational unit that independently owns assortment, pricing, and fulfillment for a set of materials. SAP-canonical Plant concept. Correct pricing grain is (material, supply_org).

Aliases: supply org, supply_org, plant, werk, werks

A price selection rule that returns one price per (material, supply_org) pair. Joins on both keys to resolve the correct price at query time.

Aliases: supply org matched, plant matched

DE: These | FR: Thèse

A testable business question evaluated against signal evidence. Theses translate technical findings into CFO-ready language. Status: confirmed, plausible, not observed, insufficient data.

Is: a testable business claim with evidence scoring

Is not: a guess or assumption. Every thesis is grounded in signal findings

Aliases: thesis, hypothesis, these, thèse, hypothese, hypothèse

DE: These-Status | FR: Statut de thèse

The per-tenant evaluation of a thesis or a Subject Matter Check (Level 1). Status: confirmed, plausible, not observed, or insufficient.

Is: a deterministic evaluation result

Is not: a finding (findings come from signals; thesis status comes from the evidence chain)

Aliases: thesis status, verdict, verdikt, urteil

DE: Behandlung | FR: Traitement

A deliberate action to change the organization’s state — process changes, staffing adjustments, policy updates. Signals detect, treatments intervene.

Aliases: treatment, behandlung, traitement, intervention

DE: Verdikt

A structured root-cause explanation for a confirmed thesis. Verdicts identify why an anomaly pattern exists, assign a root cause category, express confidence, and recommend investigation steps.

Is: a causal explanation with confidence scoring and actionable recommendations

Is not: a treatment or action — verdicts explain why, they don’t prescribe next steps (that’s the Treatment layer)

Aliases: verdict, diagnosis, verdikt, diagnose, diagnostic

Raw intake layer. Immutable copy of source system data — traceability over correctness. No interpretation, no filtering.

Aliases: bronze

DE: Kompilierung vs Build

Two distinct phases: Compile (YAML → SQL, code generation) and Build (SQL → DuckDB, per-tenant execution).

Aliases: compile, build, kompilierung, compilation

A self-contained analytical framework for a specific industry. Each pack bundles signals, perspectives, theses, verdicts, Subject Matter, dbt models, contracts, and source-system adapters.

Aliases: domain pack, paket

DE: Ephemere Datenbank | FR: Base de données éphémère

The DuckDB file (dev.duckdb) is disposable — rm + rebuild recreates everything from CSVs. The database is a derived artifact.

Aliases: ephemeral, duckdb, dev.duckdb

DE: Extraktionsvertrag | FR: Contrat d’extraction

A pipeline.yml file at the tenant AFS root that lists every file crossing the extract layer with a SHA-256 pin, source type, and expected schema. The configuration IS the contract, and the contract is the audit trail.

Aliases: pipeline.yml, extract contract, trust boundary contract

Consumption layer. Source-system-agnostic views filtering Silver to is_valid = true. This is the product contract — signals operate exclusively on Gold.

Aliases: gold

The analytical engine: CLI, compilers, validators, Explorer, and orchestration. Domain packs plug into jinflow to provide industry-specific analytics.

Aliases: jinflow, core, erp, numetrix layer

DE: Make (reine Funktion) | FR: Make (fonction pure)

The build phase as a pure function: AFS in → KLS out. No SIS mutation, no external fetches, no network calls. The security boundary is clean.

Aliases: make, jin make, jinflow make

DE: Medallion-Architektur | FR: Architecture Medallion

The Bronze → Silver → Gold → Platform layering pattern, based on the medallion architecture popularized by Databricks. Each layer has a strict responsibility: structure, validation, consumption, cross-tenant union.

Aliases: medallion, medal, medaillon, bronze silver gold

DE: Betriebsmodus | FR: Mode opérationnel

One of five deployment topologies: Local, Proxy (P2P2P), Cloud (R2), Semi-Cloud, Make-as-a-Service. Same product, different placement of AFS, KLS, SIS, and the make runner.

Aliases: deployment mode, deployment topology

Peer-to-Peer-to-People. Deployment mode where the KLS stays on the data owner’s machine and the cloud Explorer routes queries via a tunnel. Your data. Your machine. Their browser.

Aliases: peer-to-peer-to-people, proxy mode, data sovereignty

DE: Pipeline-Graph | FR: Graphe de pipeline

An interactive visualization of the entire dbt DAG — every model, source, and dependency rendered as a navigable graph.

Aliases: pipeline, graph, dag, pipeline graph

The preparation phase that gathers all inputs (source data + SIS notes/bookmarks) and commits them to the AFS so that make can run as a pure function.

Aliases: premake, jin pre-make, jinflow pre-make

DE: Wiederherstellbar | FR: Reconstructible

A table or view that can be fully recreated from source files (CSVs) and YAML definitions at any time.

Aliases: rebuildable, rebuild, wiederherstellbar

Validated facts layer. Normalized, type-cast, with is_valid flag and invalid_reason. Invalid rows are flagged, not dropped.

Aliases: silver, silber, argent

DE: Quell-Taxonomie | FR: Taxonomie des sources

The classification of the six source types that can cross the extract layer: curated spreadsheet, system export, delimited file, API pull, database query, and flat drop. Each has its own governance rules.

Aliases: six source types

DE: Quellsystem-Dispatch | FR: Dispatch système source

The macro-based mechanism that maps source-system-specific column names (OPALE, SAP, Navision) to the canonical nuMetrix schema.

Aliases: dispatch, source system, quellsystem, opale, sap, navision

DE: Souveränitätsprinzip | FR: Principe de souveraineté

Data sovereignty means the owner decides, moment by moment, whether their data is accessible. The decision is physical, not administrative. You don’t revoke a permission — you unplug.

Aliases: data sovereignty, physical access control

DE: Mandant | FR: Locataire

A hospital. Each tenant is isolated in its own DuckDB schema, defined by tenant.yml in tenants/{tenant_id}/.

Aliases: tenant, mandant, locataire, hospital, spital, hôpital

jazzisnow jinflow is a jazzisnow product
v0.45.1 · built 2026-04-17 08:14 UTC