Skip to content
Odal Node
Alpha — Odal Node is in active development. APIs, schemas, and docs may and will change before 1.0.

How the node works

The engine is the service that runs the core’s rules in production. In its simplest form it is a single node: one operator, on infrastructure they control, running the whole pipeline from raw input to a verifiable passport.

The pipeline

A node does four things, in order:

  1. Take data in — product data arrives from a file or an existing system.
  2. Validate and sign — the node checks the data against the sector’s rules and signs the result with the operator’s own key. This is where the proof-bound guarantee lives: the raw input is used to produce the proof, then discarded.
  3. Keep only the proof — the signed passport and its history are stored; the raw production data is not.
  4. Serve it — the passport is made resolvable, so a scan or a lookup returns the right view to the right audience.

The pieces

Inside the node, three surfaces handle those steps:

  • The write path — where a passport is created, validated, signed, and versioned.
  • Bulk import — for bringing many products in at once, each becoming a draft that flows into the write path.
  • The public read path — where a published passport is served and verified, and where the access tiers are enforced on every request.

Signing happens inside the node itself — the operator’s key never leaves their infrastructure. What the node keeps and what it discards is covered in Permanence & retention and Operating a node securely.

One operator per node

A node serves a single operator. There are no shared tenants and no cross-operator access; isolation is a property of the deployment, not a setting inside it. Scaling to many operators is an infrastructure concern, handled by running more nodes — not by mixing operators inside one.