Post-quantum security by design
Protect the intelligence your organisation depends on.
01
Fail-closed authority
Every content operation needs a current identity, an unexpired grant, workspace membership, a permitting role and the full handling scope, compartments and caveats included. Anything missing, malformed or expired fails closed.
02
Isolated processing
Uploaded files are treated as untrusted. Extraction runs in a separate non-root container with a read-only filesystem, capped memory and CPU, a hard deadline, and no access to database credentials or application keys.
03
Signed and recoverable
Release bundles are signed. Backups are encrypted without staging plaintext, carry a signed manifest, and restore into fresh storage only after signature, hash and image identity checks pass.
Protect national interests with post-quantum key establishment using ML-KEM, digital signatures using ML-DSA, and two independently keyed AES-256 encryption layers (dAES). Handling labels run from OFFICIAL to TOP SECRET, with versions, actors, timestamps, reasons and integrity hashes retained for documented changes.
Deployment
On-premise, private cloud or public cloud.
CORTEX ships as a signed five-service node: ingress, web application, database, isolated parser and an optional local model. Run it on your own hardware, in your private cloud, or in an AIC-operated Microsoft Azure environment in the UK. Same application, same controls, whichever route you choose.
Disconnected
On-premise or air-gapped
Browser
Local CORTEX node
Application, storage and parser. Optional local model.
External capabilities are blocked before transport. An air-gapped profile ships with the node; physical isolation and disconnected acceptance are confirmed per installation.
Connected
Private or public cloud
Browser
Local CORTEX node
Application, storage and parser. Optional local model.
Approved feeds, map tiles and hosted model endpoints only, under your organisation’s policy. Hosted in Azure UK South with independent environments, or on cloud infrastructure you operate.
AI, on your terms
Run a pinned local model on the node, or an approved endpoint you control.
01
Local model
A pinned, digest-verified model runs on your own hardware. No automatic download, no substitute model, no cloud fallback. Generation works in both Online and Offline modes.
02
Approved remote model
In Online mode, connect an OpenAI-compatible endpoint you have approved, with protected credentials and a handling ceiling. Tools, retrieval, remote storage and retries are switched off for this workflow.
03
Provenance you can inspect
Each assisted draft records the provider and model, prompt version, source and requirement versions, hashes and the first-draft fingerprint. The analyst reviews and edits before anything is saved.
Model assistance prepares an editable first draft from a bounded, authorised dataset. Review, approval and release stay with people.
Verified in build, browser and deployment.
2,506
Distinct standard-gate tests with passing evidence
105
Browser scenario executions across three configurations
175
User stories delivered across 32 feature groups
Zero
Browser errors or external requests in Offline runs
Every figure comes from named, dated runs across the application, browser suites and the release pipeline, recorded against an exact source revision. The full evidence register is available on request.
TWELVE SPECIALIST WORKFLOWS
Built for the specialist work your teams already do.
Beyond the core lifecycle, CORTEX ships the governed workflows that intelligence, policing, records, privacy and assurance teams rely on. The relevant owner switches each one on for an approved profile and purpose, under current permissions, with every decision retained.
Adopt UK scope and guidance
Approved jurisdiction, applicability and catalogues, adopted explicitly for new work.
Commission, analyse and assure
Agreed question and design, recorded confidence, structured challenge and independent assurance.
Plan and protect collection
Permitted lineage, source protection, independence and reserved capacity before tasking.
Review information and reports
Quarantine and scanning, exact-span annotation, and approved projections of received reports.
Run police tasking
Force policy, control strategy, strategic and tactical assessments, subject and problem profiles.
Handle privacy and records
Bounded processing, rights requests, retention appraisal, holds and verified disposition.
Share to an approved recipient
Recorded agreements, identification-risk review, exact projections and transfer receipts.
Explore context and futures
Adopted lenses, intelligence-estimate sections, scenarios and unintended-effect sessions.
Govern people, risks and assurance
Asset and role risk, appointments, control baselines, incident correlation and custody receipts.
Use and review AI
Registered use cases, dataset rights, evaluation thresholds, approval, suspension and retirement.
Close the learning loop
Exact-version feedback, qualified outcome evidence, assigned lessons and verified completion.
Operate and extend the node
Approved data flows, signed bundles, recovery, workload sizing and justified optional capability.
What buyers ask before they commit.
COMMON QUESTIONS
How is this different from a document repository?
A repository stores files. CORTEX keeps the requirement, the source-use authority, the exact information, the analytical judgement, the independent release decision and the recipient’s response connected to one another. Re-open an assessment years later and you see the version of the material it was based on.
Where can CORTEX run?
On your own hardware, in your private cloud, or in an AIC-operated Microsoft Azure environment in the UK. It is the same signed five-service node in every case. On-premise and private cloud installations can run an Offline policy, including an air-gapped profile, so nothing leaves the site. Public cloud runs Online with only the feeds, map gateways and model endpoints your organisation has approved.
Is it available to us, and is it ready for our operational network?
Yes on both counts. CORTEX is licensed to government departments, law enforcement agencies and private intelligence organisations. The connected application is implemented and backed by recorded application, database, browser and document-generation checks, and ships as a signed node with encrypted recovery and upgrade tooling. We then complete operational acceptance with you against your own criteria, so you know exactly what has been verified in your environment.
How does the AI work, and where does our data go?
Model assistance prepares an editable first draft from one bounded, authorised dataset, and the analyst reviews it before anything is saved. You can run a pinned local model on the node, which never calls out, or connect an OpenAI-compatible endpoint you have approved, with protected credentials and a handling ceiling. Offline policy rejects hosted generation altogether. Every draft records the model, prompt version, source versions and hashes.