Data Access Models
IUDX supports a range of data flows so consumers can access data in the way that best fits their use case. Every model is governed by the same consent and access-policy fabric, and every access is authenticated and logged.
1. Bulk Download
A copy of a dataset is downloaded to authorized users.
- Best for: static or historical datasets, and large-scale offline analysis.
- Enables: integration into the consumer's own systems.
2. Pull-based API Access
Regulated, policy-based, on-demand access on a per-record basis through secure, standardized APIs (harmonized with NGSI-LD).
- Best for: controlled, metered consumption of live or frequently updated data.
- Enables: fine-grained, auditable access aligned to consent and policy.
3. Push-based Streams
Real-time updates streamed to subscribed consumers via the data broker.
- Best for: IoT and time-series data, and applications that react to live events.
- Enables: low-latency, event-driven integration.
Choosing a model
| Model | Pattern | Typical use |
|---|---|---|
| Bulk Download | Full copy to user | Static / historical datasets, large-scale analysis |
| Pull-based API | On-demand, per-record | Controlled, metered access to live data |
| Push-based Stream | Real-time subscription | IoT, time-series, event-driven applications |
Data classes
Access also depends on a dataset's class. The platform supports public data (openly accessible), restricted data (access granted by the provider on request, per policy), and monetisable data models (sold through the Marketplace). Restricted access is governed by the provider's policies in the Policy Service.
:::note Emerging: privacy-preserving compute For sensitive data, CDPG is developing privacy-enhancing techniques — including computation inside secure enclaves so that analysis can run without the raw data leaving the provider's control. These build on the same consent and policy foundation. :::