USE CASES · CORTEX
Intelligence software for intelligence organisations
Produce defensible intelligence assessments for national-security decisions.
CORTEX gives intelligence organisations one connected workflow for requirements, collection, analysis, dissemination and feedback. Analysts can trace an assessment back to its sources, test competing explanations and prepare a reviewed product for named recipients. JDP 2-00-informed workflows and PHIA probability terminology support professional intelligence production, while post-quantum protection helps safeguard sensitive reporting and national interests.
National-security assessment workflow
Example workflow: assess an emerging threat before a policy or operational decision. Record the intelligence requirement and deadline, check existing holdings through an RFI, then task authorised collection to fill the remaining gaps. Compare hypotheses using cited evidence, record probability and confidence separately, and submit the assessment for independent review. Release the approved version to named recipients and use their feedback to decide the next requirement.
01 · Intelligence requirements and priority setting
Give each intelligence requirement a question, decision point, priority, deadline and accountable owner. CORTEX supports intelligence requirements (IRs), priority intelligence requirements (PIRs), specific intelligence requirements (SIRs) and essential elements of information (EEIs), with linked subordinate needs and retained revisions.
JDP 2-00-informed terminology helps teams organise direction consistently. Changes to scope, ownership and status retain their reasons, so a new analyst or commissioner can understand why the work was commissioned and what remains open.
02 · RFI management and holdings review
Answer requests for information from existing holdings before commissioning additional collection. An eligible internal responder can return cited evidence or record a nil, refusal or referral outcome. The answer remains attached to the request and the material used to address it.
Keep holdings review and new acquisition as distinct decisions. Where an RFI exposes a gap, raise the appropriate requirement and collection task rather than treating an incomplete answer as a completed intelligence need.
03 · Source protection and collection management
Record source provenance, reliability, restrictions and permitted use before tasking collection. Plans connect the approved source revision and intelligence requirement to a collector and a UTC task window. Results capture observations, credibility assessments and outstanding gaps for each attempt.
Assess source reliability separately from information credibility. Retain common reporting origins when checking corroboration, and preserve historical results when source authority changes or a task is repeated.
04 · PHIA-informed analysis and independent review
Use PHIA probability terminology alongside a separate confidence assessment and supporting rationale. Analysts can record hypotheses, assumptions, contradictory evidence and significant disagreement. Competing-hypotheses matrices and time-bounded judgements support structured analytical challenge.
Assurance plans record findings against the eight PHIA analytical standards and distinguish verification from fitness for the decision. An eligible noncontributor reviews the submitted revision; changed reasoning returns to draft for renewed review.
05 · Controlled intelligence reporting and dissemination
Turn a reviewed analysis into a Brief, Assessment or Intelligence Assessment with exact citations, an information-as-of date and clear limitations. Export the saved revision as PDF, DOCX or JSON, then take it through the separate release review, approval and publication workflow.
Approval applies to the content and intended internal audience. Changes to either require the appropriate renewed decision. This gives reviewers and publishers a clear record of what was authorised, while keeping receipt separate from publication.
06 · Recipient feedback and requirement review
Connect recipient feedback to the exact released package. Named recipients verify and acknowledge the package before recording usefulness, timeliness, corrections and remaining information needs. Analysts and requirement owners can then see which questions still need work.
Close, reopen or retask the requirement through an accountable owner decision. Earlier reporting, acknowledgement and feedback remain available for later review, giving the next intelligence cycle a documented starting point.
07 · Analyst-controlled AI drafting
Use an approved model to prepare a first draft from a selected, ready CSV extraction linked to the requirement. The analyst reviews the output, edits the report and submits it through the normal assurance and release process.
Approved purposes, permitted inputs, provider configuration and human review define the scope of assistance. Source authority, citations and independent approval remain part of the workflow, including when AI helps with drafting.
08 · Post-quantum security for national interests
Protect intelligence with ML-KEM key establishment, ML-DSA signatures and independently keyed double AES-256 (dAES). Post-quantum cryptography addresses evolving threats to confidentiality and integrity, including the long-term sensitivity of national-security information.
Plan identity, key custody, revocation and recovery for the deployment. JDP-informed practice and post-quantum protection support the security case; accreditation and acceptance remain decisions for the organisation and its operating environment.
Example workflows illustrate the documented capabilities. Deployment acceptance, lawful use and accreditation are assessed for each organisation and environment.