The jinflow

The product, said in one sentence.

Jinflow fosters Progress by gaining Understanding through Observations, Knowledge, and Data.

Five names. One shape. One centre. Drop any one of the five and the shape collapses — the model is held by all five.

Vocabulary

Every analytical move has the same shape.

Five names — not layers, not pillars, not aspects. The product's shape is the relationship between them.

Observations
What we notice
top-down — signed notices by named people, with state

Knowledge
What we know
pack-shipped wisdom + activatable patterns; tenant adds 50cents

Data
What the data says
bottom-up — produced by signals + their dbt rules

Understanding
The resolution
findings (atomic) + verdicts (aggregate) — the shared currency

Progress
What we'll do
downstream — suggestions, scenarios, interventions, direction
The shape

Understanding at the centre. Three feeders converging.

Observations
Knowledge
Understanding
Data
Progress

Observations enter from above. Knowledge informs from one side. Data feeds in from the other. Progress flows out below. Understanding is the only place where all three feeders agree on what's true. The three feeders develop in parallel — never sequentially.

What holds it together

Three feeders. One resolution. One forward arm.

Observations

A signed notice by a named person — "I suspect we're losing margin on generic substitutions." It enters with stake, carries a state, seeks ratification from Understanding.

Knowledge

What the system already holds before Data runs. Wisdom from the pack, 50cents from the tenant, and the library of activatable patterns the pack ships. It informs what Understanding can defensibly say at every layer.

Data

What the data says. Bottom-up: Signals scan, Theses ask, Verdicts judge. Declarative, typed, compiled to SQL. The empirical ground every notice gets tested against.

All three converge at Understanding. From there, Progress takes validated truth into modelled action.

Data · what the data says

Three tiers. Each producing a distinct artifact.

Asks
Signal

"What can we detect?"

Many rows per tenant — one per matching entity / time bucket. Lands as signal_findings__<id>.

Asks
Thesis

"Is this happening?"

Exactly one row per (tenant × thesis). status ∈ {confirmed, plausible, not_observed, insufficient}.

Asks
Verdict

"Why is this happening?"

Many rows per tenant — one per matched root-cause case. Lands as verdict_findings.

Theses and verdicts are pack-shipped Knowledge patterns; their dbt models run inside Data and emit into Understanding.

Staged disclosure

The data has to earn its narrative, one tier at a time.

Thesis gate

The thesis a verdict is bound to must be confirmed. If plausible, not_observed, or insufficient, the verdict never fires — even if signals scream.

Condition gate

The verdict's own conditions — signal references with field thresholds — must match. The audience sees facts before interpretation, interpretation before explanation.

A thesis that doesn't confirm acts as a silencer for the verdict layer below it. The result: no premature explanation.

Observations · what we notice

An entity. With state. With facets.

An Observation is the entity. Explanation and Contributing Factor aren't peers — they're facets the Observation carries.

Entity
Observation

A signed notice about a specific situation. Authored by a human. State: suspected → validated / refuted / inconclusive → resolved.

Facet
Explanation

"Why is the observation true?"

A facet the Observation carries. Composed of one or more Contributing Factors.

Facet
Contributing Factor

"Which Understanding backs this?"

A citation of a specific Finding or Verdict. Granular and individually-citable.

Understanding · the resolution

Where evidence becomes narrative.

Findings

Atomic — one row, one observation. Each carries severity, entity, time, money at risk, machine-readable evidence. The unit of detection.

Verdicts

Aggregate — one row per (tenant × thesis), summarising whether the thesis holds. The unit of judgment.

Together they are the shared currency. Observations ratify against Understanding. Data lands its detections here. Knowledge informs how every row should be read. Progress draws its starting point from here.

Knowledge · what the system holds

Subject Matter. Plus activatable patterns.

Two kinds of inhabitants. Both pre-formed before Data ever runs. Each carries a provenance state — pack-original, pack-modified, or tenant-original.

Subject Matter · always-on

Wisdom from the pack. 50cents from the tenant. Dossiers group related entries into reading guides. Atomic expert knowledge that informs Understanding without being produced by it. "Lehrauftrag positions without recent operational activity are normal here, not a ghost employee."

Activatable patterns · opt-in

Theses (question-shape) and Verdicts (root-cause-shape) shipped pre-formed by the pack. The tenant engages the ones that match its situation — the rest sit on the shelf. The pack's analytical lens collection.

Progress · what we'll do

Where validated truth turns into change.

Steers
Direction

The attitude, goals, intent that suggestions are validated against. "More employee happiness, or less salary spend?"

Surface
Suggestion → Scenario

Suggestion → Intervention → Scenario → Simulation. Each step a tightening: "what could be done""what the delta actually measures."

Surface
Governance moves

Interventions in the world — update a contract, file a credential renewal, stage a hiring decision. Don't live in the data, but produce future deliveries the Data side will see.

Progress is the last stop in the diagram and the first input to the next cycle.

What the model prevents

Three things analytical platforms historically conflate.

No premature explanation

Without thesis confirmation, verdict findings never fire. The audience never sees root-cause speculation about theses the data doesn't yet support. Understanding is staged.

No unfounded narrative

Without analytical ratification, an Observation can't progress past suspected. A hypothesis stays a hypothesis until evidence arrives. Observations ratify against Data.

No siloed knowledge

Knowledge sits as a peer feeder of Understanding and is queryable from any of the four other names. Subject Matter modifies what verdicts can defensibly say AND what observations can defensibly say. Knowledge that lives only in someone's head doesn't.

The whole product

In one sentence.

Jinflow fosters Progress by gaining Understanding through Observations, Knowledge, and Data.

A jinflow pack ships the rules for all five names. A jinflow tenant produces the rows those rules emit when run against its data. The narrative on top is authored by humans who care about the outcomes — and is verifiable against the analytical evidence underneath.

Observations
Knowledge
Understanding
Data
Progress