01
Sources
Registries, open web, dark-web context, regional records, customer context, and selectors.
Capabilities
Capabilities for building the intelligence layer.
Stratir connects sources, ontology, agents, analyst review, and delivery into operating layers that help teams make better use of scattered information.
SYSTEM DESIGN SURFACE
Requirement, source context, entity logic, and review gates stay connected from intake to delivery.
Source to decision
REVIEWABLE OPERATING MODEL01
REQUIREMENT02
SOURCE MAP03
REVIEW LOOP02
Ontology
People, companies, documents, events, sources, and findings become connected objects.
03
Agents
Bounded agents help collect, enrich, compare, summarize, test, and QA the work.
04
Review
Analysts validate uncertainty, provenance, source quality, and what should be delivered.
05
Delivery
The output can be a brief, case file, lead package, tool, operating layer, or training path.
Six lanes, one operating layer.
The point is not to add another dashboard. The point is to make the work clearer: what was found, where it came from, how it connects, and what still needs human judgment.
Source layer
01
All-source data aggregation
Unify selectors, registries, business records, open-source traces, identity signals, and dark-web context into one structured investigation surface.
Feynman proof of work, CPF/CNPJ pivots, KYB registry paths, AML research, selector lookup.
Model layer
02
Ontology-first workflows
Represent people, companies, selectors, documents, regions, and evidence as connected objects rather than flat search results.
Entity resolution, relationship mapping, case memory, source history.
Context layer
03
Regional intelligence coverage
Design investigations around geography, jurisdiction, language, and source availability, especially where data is fragmented or operationally uneven.
LATAM, Singapore, Hong Kong, trade finance, registry-aware operating contexts.
Control layer
04
Clandestine research and OPSEC
Support lower-noise research where controlled pivots, source handling, and sensitive workflows matter as much as speed.
Quiet-mode research, OPSEC posture, controlled collection discipline.
Operating Cloud
One layer for the moving parts.
Sources, objects, agents, and review gates stay connected from question to delivery.
SourcesOntologyAgentsReview
Execution layer
05
Agentic engineering
Use agents as an execution layer for research, engineering, enrichment, QA, and synthesis without losing human control.
Codex, Cursor, Hermes, Pi.dev, security review, local verification loops.
Transfer layer
06
Training and tradecraft transfer
Turn intelligence methodology into structured education for analysts, investigators, and teams that need repeatable tradecraft.
OSINT, CTI, DarkNet, OPSEC, multilingual learning paths.
The operation decides where the tool belongs.
Stratir starts with the decision and works backward into sources, models, agents, review, and delivery.
01
Question
Define the business question, decision owner, sensitivity level, and acceptable evidence standard.
02
Collection
Acquire relevant public, registry, regional, and source-contextual data without indiscriminate scraping.
03
Processing
Normalize selectors, entities, dates, jurisdictions, documents, and source metadata into a usable model.
04
Analysis
Correlate evidence, test hypotheses, preserve uncertainty, and keep the analyst in control.
05
Delivery
Deliver briefs, evidence packages, graphs, leads, tools, and operating layers another professional can inspect.
Useful when the answer needs evidence behind it.
Due diligence, KYB, and entity research.
AML investigations and counterparty review.
Fraud research, impersonation mapping, and case-building.
Dark-web monitoring and early risk signals.
Corporate security, vendor risk, and executive protection support.
OPSEC-aware research and AI-assisted security engineering.