DataFab OSINT Service

Version: 1.0 Last Updated: September 2026

Component Overview

The OSINT service is an external enrichment component of the DataFab platform. On request from the client-side Knowledge Fabric, it performs open-source intelligence (OSINT) lookups: it searches publicly available sources (public web, news and corporate registries) for an entity the client names, and returns the findings in a structured form. It performs no profiling, scoring or decision-making of its own; it collects and organises publicly available information and hands it back.

The service is the external counterpart of the Knowledge Fabric’s client-side enrichment integration (see External Source Integration). The client side owns the request schema, source preferences, result fusion and storage; the OSINT service owns only search execution and structured extraction.

Knowledge Fabric (customer env) ──outbound TLS──▶ OSINT Service (DataFab AWS)
                              ◀──structured results with source URLs──

Deployment and Connectivity

Aspect	Position
Hosting	Deployed in the DataFab AWS environment, in a region agreed with the client
Direction	One-directional — the client environment opens outbound TLS connections to the service API
Inbound access	None — the service never initiates connections into the client environment and holds no credentials for it
Authentication	Per-tenant API key issued to the client
Limits	Requests are rate-limited and budget-limited per tenant

What Data the Service Receives

The service receives only what the client-side integration sends, and the shape of that input is fixed by the request schema defined and controlled on the client side. In practice this is the minimum needed to run a search: an entity name, an entity type (person or organisation), and optionally a small set of disambiguating attributes and the list of fields the client wants populated.

The service has no access to client systems, case files, client records or any data beyond the request body.

Request Element	Required	Controlled By
Entity name	Yes	Client-side request schema
Entity type (person / organisation)	Yes	Client-side request schema
Disambiguating attributes	Optional, minimal set	Client-side request schema
Fields to populate	Optional	Client-side request schema

What Data the Service Produces

All returned information is collected from public sources. Each returned value carries the URL of the public source it was taken from, so the client can verify provenance. The result is returned to the client service in the API response; the client decides whether and where it is stored.

Result Property	Description
Content	Publicly available information only
Structure	Conforms to the field list requested by the client
Provenance	Every value carries the URL of its public source
Delivery	API response only — the client controls storage

Retention

The service acts on a request-and-return basis. Request inputs and results are held only for the duration of the job so the result can be delivered and the call can be metered, and are then purged. No copy of the request or result is retained for any other purpose.

Long-term records on the DataFab side are limited to non-content operational metadata, used for rate limiting, billing and audit. These records contain no entity names, attributes or results.

Data	Retention	Purpose
Request inputs (names, attributes)	Duration of the job, then purged	Search execution
Results	Duration of the job, then purged	Delivery to client
Operational metadata (tenant ID, endpoint, timestamp, HTTP status, cost units)	Long-term	Rate limiting, billing, audit