Agentic OSINT
Threat Intelligence
A multi-agent threat intelligence platform purpose-built for critical infrastructure operators. Continuous monitoring across four security domains — delivered continuously.
Overview
AGENTIQ Security OSINT is a multi-agent threat intelligence platform purpose-built for critical infrastructure operators. It continuously monitors open-source intelligence across four security domains — Cyber, Physical, Personnel, and Supply Chain — and delivers structured intelligence reports without human intervention.
Designed as a systems-integration product, AGENTIQ is customised for each client’s sector, geography, and threat landscape. Organisations can self-host the platform on their own infrastructure under a CAPEX ownership model, or subscribe to a fully managed OPEX service with dedicated support.
The Problem
Critical infrastructure operators face an asymmetric threat landscape. Adversaries span nation-states, cybercriminals, insider threats, and physical saboteurs — yet most security teams rely on fragmented, manually curated intelligence that covers only one domain at a time.
Traditional OSINT solutions are expensive, require dedicated analyst headcount, and rarely deliver cross-domain correlation. Smaller operators — regional utilities, water authorities, transport networks — are left with generic threat feeds that generate noise, not insight.
What AGENTIQ Does
Automated Collection
AGENTIQ deploys a fleet of specialised AI agents that continuously collect, clean, and categorise data from curated open-source feeds. Out of the box, the platform monitors 16+ sources including government advisories, vulnerability databases, and global news — with sources tailored to each client’s sector and geography.
Collection runs daily with no AI compute cost, using a proprietary cleaning pipeline that typically strips 99%+ of raw noise from HTML, XML, JSON, and RSS data before any AI processing occurs.
Cross-Domain Analysis
A dedicated Analyst agent applies structured intelligence methodologies to identify Tactics, Techniques, and Procedures (TTPs) across all four security domains:
| Domain | Example Coverage |
|---|---|
| Cyber | Vulnerability exploitation, ransomware campaigns, SCADA/ICS threats |
| Physical | Substation sabotage, copper theft, terrorism targeting infrastructure |
| Personnel | Insider threats, social engineering, credential harvesting |
| Supply Chain | Software supply chain compromise, vendor risk, dependency attacks |
Every TTP is graded using the NATO Admiralty System for source reliability and information credibility, giving decision-makers a standardised confidence framework.
Deterministic Reporting
AGENTIQ produces two flagship reports:
- TTP Intelligence Report — A structured, technical assessment for security teams, featuring individually scored threat cards mapped to the MITRE ATT&CK framework where applicable.
- Executive Intelligence Summary (EIS) — A concise, non-technical brief for leadership and boards, including domain-level threat summaries, convergence scenarios, and prioritised monitoring recommendations.
Reports are generated as branded, paginated PDFs using a deterministic JSON-to-template architecture. This means formatting is guaranteed — no layout drift, no missing sections, no surprises.
Source Provenance & Anti-Hallucination
AGENTIQ implements a three-layer provenance system that ensures every cited source in a report traces back to a verified data feed:
- Source Registry — Only pre-approved, real-world sources can be cited
- Prompt Constraints — The AI is explicitly prohibited from fabricating or inferring sources
- Post-Processing Validation — Every source URL is programmatically verified against the registry before publication
The result: no hallucinated sources observed across all production reports to date.
Flash Alerts
Between weekly analysis cycles, AGENTIQ scans daily collections for critical threat patterns — such as active exploitation of critical infrastructure systems — and can trigger real-time email alerts to nominated stakeholders when high-confidence matches are detected.
Architecture
AGENTIQ is built on a multi-agent orchestration model with four specialised roles:
| Agent | Function |
|---|---|
| Collector | Gathers and categorises raw OSINT from curated feeds |
| Analyst | Applies structured methodology to identify and grade TTPs |
| Reporter | Synthesises analysis into standardised intelligence products |
| Orchestrator | Coordinates the pipeline, manages sequencing and error handling |
The platform runs inside a hardened, sandboxed environment with strict network isolation policies. Only explicitly whitelisted endpoints are reachable — all other traffic is denied by default. This architecture is designed for environments where security, auditability, and data sovereignty are non-negotiable.
Hybrid Execution Model
AGENTIQ separates collection from analysis to optimise cost and performance:
- Daily — Automated data collection and cleaning across all sources. Zero AI compute cost.
- Weekly — AI-powered analysis, TTP identification, report generation, and delivery. Current typical cost per cycle: under $0.20 (subject to model pricing).
This hybrid approach can significantly reduce operational AI costs compared to daily full-analysis models, while maintaining comprehensive threat coverage.
Deployment Models
AGENTIQ is a systems-integration product, customised for each client’s operational environment.
CAPEX — Self-Hosted Ownership
The platform is deployed on the client’s own infrastructure — on-premises, private cloud, or air-gapped environments. The client owns the system outright.
- Full source code transparency and audit rights
- Custom source selection and reporting cadence
- Optional support and maintenance agreements
- Ideal for organisations with sovereign data requirements or existing security operations centres
OPEX — Managed Subscription
AGENTIQ hosts, operates, and maintains the platform on behalf of the client. Reports are delivered to nominated stakeholders on a defined schedule.
- No infrastructure overhead
- Includes ongoing source curation and platform updates
- Dedicated support
- Ideal for organisations that want intelligence outcomes without operational complexity
Current Capabilities (v0.3)
| Capability | Status |
|---|---|
| Automated daily OSINT collection (16 sources) | ✅ Production |
| Cross-domain TTP analysis (4 domains) | ✅ Production |
| NATO Admiralty grading | ✅ Production |
| Branded PDF reporting (TTP + EIS) | ✅ Production |
| Deterministic JSON/Jinja2 report architecture | ✅ Production |
| Three-layer anti-hallucination provenance | ✅ Production |
| Flash Alert pattern scanning | ✅ Production |
| Hybrid execution model (daily collect / weekly analyse) | ✅ Production |
| Automated email delivery to stakeholders | ✅ Production |
| Hardened sandbox with network isolation | ✅ Production |
| Automated cron scheduling | ✅ Production |
Roadmap
Near-Term (v0.4–v0.5)
- Longitudinal trend tracking — SQLite-backed historical analysis to identify emerging threat patterns over time
- Automated instance cloning — Single-script deployment for rapid client onboarding
- Per-client branding and source presets — Multi-tenant configuration management
- Interactive dashboard — Real-time visibility into collection status, source health, and threat trends
Medium-Term (v1.0)
- Sector-specific intelligence packs — Pre-configured source sets and analytical frameworks for energy, water, transport, telecommunications, and healthcare
- Bi-directional API integration — Ingest from and push to existing SIEM, SOAR, and GRC platforms
- Collaborative intelligence — Cross-client anonymised threat sharing for sector-wide situational awareness (opt-in)
- Regulatory mapping — Automated alignment of identified threats to compliance frameworks (SOCI Act, NIST CSF, ISO 27001, AESCSF)
Long-Term Vision
- Predictive threat modelling — AI-driven forecasting of threat actor behaviour based on historical pattern analysis
- AI-driven response recommendations — Context-aware mitigation strategies tailored to the client’s operational environment
- Continuous real-time monitoring — Transition from periodic collection to persistent, event-driven intelligence
- Edge deployment — Lightweight collection agents deployable on operational technology networks for direct ICS/SCADA visibility
Why AGENTIQ
- Cross-domain by design — Many platforms focus primarily on cyber threats. AGENTIQ analyses physical, personnel, and supply chain threats in the same pipeline.
- Rigorous anti-hallucination controls — Every source is verified against a pre-approved registry. Every report is auditable.
- Built for critical infrastructure — Hardened, isolatable, deployable in sovereign and air-gapped environments.
- Structured intelligence output, continuous operation — Reports modelled on established intelligence reporting standards, generated without human intervention.
- Transparent economics — No per-seat licensing. No query limits under current plans. Predictable costs whether self-hosted or managed.
Sectors
AGENTIQ is designed for any organisation responsible for critical infrastructure, including but not limited to:
- Electricity generation, transmission, and distribution
- Water and wastewater utilities
- Oil, gas, and pipeline operators
- Transport and logistics networks
- Telecommunications providers
- Healthcare and hospital networks
- Government and defence facilities
- Ports and maritime infrastructure
Get Started
AGENTIQ Security OSINT is available for deployment now. Contact us to discuss your sector, threat landscape, and preferred delivery model.
Get in Touch