jinflow

Your data. Your machine. Their browser.

An analytical engine that makes insight portable, sovereign, and durable. The data never has to leave the laptop that produced it.

Today

Dashboards are opinions. Spreadsheets die. Knowledge gets locked in slide decks.

Dashboards lie quietly

They show what was queried, not what's missing. The interesting findings are the ones nobody wrote a tile for.

Sensitive data sleeps in pipes

Cloud BI ingests everything to render anything. Compliance teams spend years saying no.

Expertise evaporates

The analyst who knows that OPALE_ABT_42 means surgery leaves. Their replacement starts at zero.

What jinflow does differently

Declarative. Transparent. Independent. Human.

Declarative

Signals, theses, verdicts as YAML. Compiled to SQL. Validated at every layer.

Transparent

Every finding traces back to its source row. Every verdict explains its evidence.

Independent

The Knowledge Store is one DuckDB file. Runs on a laptop. No cloud required.

Human

Subject Matter captures expert knowledge that no model can derive — and travels with the build.

How it works

AFS in. KLS out. One make.

AFS — Artifact Store

Git-versioned YAML. Signals detect anomalies, theses ask business questions, verdicts explain why. Validated end-to-end before any SQL runs.

KLS — Knowledge Store

One DuckDB file. Signed snapshots, self-describing, immutable. The build product travels with its source.

Five operating modes

Local, peer-to-peer-to-people, cloud, semi-cloud, make-as-a-service. Same code converges. Sovereignty by construction.

Live, today

One engine. Many domains. Same shape, different worlds.

Hospitals, logistics, winemaking, alpine resorts, legal practices — each runs on the same engine through a domain pack that adds the vocabulary. HR is forming next. Yours could be among the next.

Now

Built on dbt + DuckDB. Bring your domain.

jinflow runs anywhere a laptop runs. The engine is open. Domain packs are starter kits — yours, or one of ours.

Michael Gassmann · jazzisnow
hello@jinflow.io