Four roles in one software: how a single connector can be provider, consumer, broker, and clearing house

Most data space architectures assign a fixed role to each component: the provider is a provider, the consumer is a consumer, and if you need a broker, you deploy entirely different software. Our data space breaks this rigidity: a single connector can simultaneously operate in four distinct modes, adapting to the role each organization needs to play at any given time.

Provider: sharing data with the world

In provider mode, the connector creates local catalogs, publishes datasets with metadata and usage policies, and exposes them to other data space participants. When a remote consumer requests access, the provider's connector manages the negotiation, formalizes agreements, and serves data through its Data Transfer Module.

The provider maintains total control over their data: they define who can access it, under what conditions, for how long, and can suspend or terminate any transfer at any time. Data never leaves their infrastructure without passing through policy and agreement validations.

Consumer: search, negotiate, and obtain data

In consumer mode, the connector browses remote catalogs (directly from providers or through brokers), initiates negotiations to access datasets of interest, and once an agreement is reached, requests transfers to access the data.

The interesting part is that received data isn't a dead end: the consumer can convert any active transfer into an internal data source and connect it to processing pipelines. This allows, for example, receiving air quality data from a provider, passing it through an analysis application to detect anomalies, and storing results in their own database.

Metadata broker: the data space hub

Broker mode transforms the connector into a catalog aggregator. Instead of forcing each consumer to know every provider's address, the broker collects and synchronizes catalogs from multiple providers into a single query point.

Enabling broker mode is as simple as toggling an option in the data space configuration. Once activated, providers can request registration with the broker (presenting a verifiable identity and name attestation), and the broker administrator reviews and approves requests. The broker periodically syncs registered providers' catalogs, keeping information current.

For consumers, querying a broker is identical to querying a provider: the search and discovery experience is the same, but the catalog spans the entire data space.

Clearing house: the data space's digital notary

Clearing house mode adds the ability to act as a trusted third party. When other participants negotiate agreements, they can designate this connector as the clearing house to register both parties' digital signatures, store agreements, and record them on blockchain.

The clearing house also manages a dispute system: if provider and consumer disagree about whether conditions were met, they can exchange dispute messages through the clearing house, which acts as an impartial mediator with access to stored cryptographic evidence.

And just like broker mode, enabling the clearing house doesn't require deploying additional software: it's activated from the existing connector's configuration.

Simultaneous modes: real flexibility

The key to this architecture is that modes are not mutually exclusive. An organization can simultaneously be a provider of its own data, a consumer of third-party data, and operate a broker to facilitate discovery within its consortium. A regulatory body could be both a clearing house and a reference data provider.

This dramatically reduces infrastructure costs and operational complexity. Instead of managing four different software pieces, one is deployed and maintained, with configuration adapted to the needed role at any given time.

One connector, infinite configurations

The "one connector, multiple roles" philosophy is not just an architectural choice: it reflects the reality that European organizations don't fit into rigid categories. A university may provide research data, consume environmental data from public administration, and operate a broker for its partner network. With our data space, all of this is managed from the same software, with the same interface, without duplicating infrastructure.

Fewer servers, less maintenance, more participants: that's the equation a truly versatile connector solves.