INTELLIGENCE DATA FABRIC
NERVE
From source to searchable information.
NERVE is the ingestion, preservation, transformation and search layer beneath CORTEX. It connects approved sources to a governed data lake, keeps the original intact, normalises source-specific structures into one common information model and exposes authorised search with exact evidence references.
Product Type
Intelligence Data Fabric
Product Code
AIC-NRV
Deployment
NERVE Local / Disconnected / Private Cloud / Approved Cloud, exposing the same logical contract across each profile
Key Differentiator
One governed ingestion layer that preserves originals and retains provenance, so every approved source becomes searchable in CORTEX without a new point-to-point integration.
The Ingestion Layer Beneath CORTEX
NERVE is AIC’s intelligence data fabric. It acquires information from approved digital sources, preserves the exact original, transforms source-specific structures into a common information model, enriches and indexes the result, and exposes governed search and retrieval to CORTEX and approved integrations.
The design goal is simple to state. The origin of a record should be irrelevant to how an analyst discovers it, while remaining fully visible for evidence and assurance. A message from an authorised device extraction, an email from a mailbox, a row from a CSV, a transaction from a database, an observation from a geospatial feed and a document from a file share should all become searchable through consistent concepts, without losing the source-specific detail that may matter later.
NERVE owns the technical path from source to searchable information: acquire, preserve, inspect, decode, normalise, enrich, index, serve. CORTEX owns the intelligence lifecycle around that information: direction, collection governance, analysis, product, dissemination and feedback. NERVE gets information in. CORTEX turns information into intelligence.
The Challenge
Intelligence and investigative teams rarely suffer from a shortage of data. They suffer from data arriving in incompatible structures: device extractions, office files, messages, data exports, application databases, cloud systems, sensors, feeds, geospatial files and legacy repositories.
Each source brings its own schema, access method and provenance story. Teams spend significant effort locating data, learning proprietary formats, writing one-off importers and rebuilding provenance every time information moves between tools. Analysts end up learning source systems instead of pursuing intelligence questions, and the path from a finished judgement back to the original material becomes difficult to reconstruct.
The Capability
NERVE is designed to solve that infrastructure problem once. Each approved source is acquired through a versioned connector or import path. Exact source material is preserved before any transformation takes place. Versioned processing runs then turn source-specific structures into canonical messages, documents, calls, contacts, locations, files, transactions and other common records, while retaining source-specific detail and full lineage.
Search is a deliberate publication state rather than a side effect of storage. A record becomes discoverable once it carries a complete minimum envelope, and every result resolves to a stable reference for the exact object revision plus the locator needed to recover the matched passage, row or cell.
Optional enrichment such as OCR, transcription, translation and embedding generation runs as separate, provenance-tracked derivative stages. Machine-derived content improves discovery without ever replacing source-derived content or standing in for analyst judgement.
Operational Outcomes
Integrate a source once, then reuse it across analyst workflows instead of rebuilding point-to-point imports.
Preserve exact original material so content can be reprocessed later when better parsers or models arrive.
Search one concept, such as an identifier, time window or location, across records that arrived from very different systems.
Retain provenance through transformation, so a search hit leads back to the acquisition record and the original bytes.
Give CORTEX analysts governed discovery inside the lifecycle tool rather than access to every underlying source system.
Keep stable evidence references valid after reprocessing, so earlier citations continue to resolve.
Operate the same logical contract on a disconnected node, in private infrastructure or in approved cloud.
Authority, Handling and Provenance
Access control is designed into every read path. Search is authorised before matching, counting or aggregation, so denied records do not leak through result counts or facets. Handling markings propagate conservatively, and derivatives produced from several inputs inherit the union of their access restrictions.
CORTEX integration is service-mediated. Consumers reach NERVE through authenticated APIs or approved event contracts rather than querying the object store, metadata database, search index or broker directly. Each analyst search carries both the CORTEX workload identity and the end-user authority context, so a trusted service cannot use its own technical privilege to widen an analyst’s information scope.
Security classification, privacy sensitivity and legal or policy authority are treated as separate dimensions. Retention runs by policy and hold, deletion is verifiable and scoped, and audit records are append-oriented and content-minimised. Vector indexes inherit the same security and privacy envelope as the records they describe.
Capabilities
Source and connector catalogue; bulk, API, feed and streaming ingestion; device and forensic package import; immutable original preservation; hashing and acquisition records; versioned transformation runs; common information model with source-specific extensions; provenance graph; OCR, transcription, translation and embedding enrichment as tracked derivatives; governed search and faceting; exact evidence references; retention, holds and verifiable deletion; append-oriented audit; NERVE Explorer and operator surfaces
Integrations
CORTEX search and evidence APIs; AIC.Foundation and AIC.Foundry identity and platform services; S3 and Azure Blob compatible object storage; message-driven connectors; enterprise repositories, databases, APIs and feeds; forensic tool export packages; approved OCR, transcription, translation and embedding providers
Identity Support
Human access through the approved AIC.Foundation and AIC.Foundry identity path; service access through short-lived workload identity with user authority delegation for analyst-facing reads
Data Residency
Customer-controlled. Designed for local, disconnected, private and approved cloud deployment without mandatory public cloud dependency.
Encryption
Policy-controlled signing, hashing and encryption with TLS and mTLS support; algorithm selection follows the approved deployment profile rather than being embedded in business logic.
