NERVE is live across local, private and approved-cloud deployments. Book an evaluation
Audience
Integration engineers and analysts working with partner or internal APIs.
Source families
APIs and feeds
The workflow
01
Input
An approved API or feed, a credential that NERVE is authorised to use, and a raw-preservation policy that states what is kept and for how long.
02
Acquisition and preservation
Responses and events are retained as received, together with the source schema version that described them. Pagination state is checkpointed durably so a run always knows exactly where it reached.
03
Transformation
Typed canonical records are produced from the retained payloads. Fields the canonical families do not cover remain available as namespaced extensions rather than being discarded.
04
Discovery and retrieval
Feed records are searchable with the same envelope fields as every other source. A result cites the exact retained response it came from, including the schema version in force at the time.
05
Governance
Egress is policy-controlled. A feed being readable does not make it exportable: export rules are evaluated separately from acquisition rules.
When it goes wrong
Failure variant
The run crashes mid-page. It resumes from the last durable checkpoint, and stage outputs are not multiplied. A replayed page produces the same canonical records rather than duplicates.
What this does not claim
Retained payloads reflect what the source returned at that moment. A later change at the source does not alter what NERVE already holds.
Evaluation measures
Replay handling without duplicate canonical output
Checkpoint correctness after interruption
Policy-controlled egress behaviour
Schema-version fidelity on retained payloads
Source-to-search time for new events