AUTOPURPLE Request Evaluation
Active Defense · Security Automation

Red-team insight. Blue-team discipline.
Unified as controlled autonomy.

AutoPurple builds active-defense and security-automation products on a legal, controlled, auditable foundation — led by two flagship platforms, TALON (threat deception & attribution) and SAOAS (autonomous security decision orchestration), plus a four-product suite. Every capability traces back to a systematic body of adversarial security research.

Purple = Red ∩ Blue One company, one engineer Private deployment Evidence-driven
Contact Engineer & Officer for Evaluation →
AutoPurple — Security Technology Corp.
The attacker's view, delivered as the defender's certainty
Flagship Platforms

See the adversary, act decisively — closed into one loop

Two flagship lines complete each other: TALON profiles the adversary and traces the source inside the defender's own deception environment; SAOAS takes that high-confidence intelligence as input for controlled autonomous decision and response.

TALON
Threat Deception & Attribution — the intelligence source and attribution hub of active defense
Legal, never out-of-bounds Passive harvesting Operator profiling Tiered source convergence Confidence & physical bounds
AP-WP-TALON-2026-001 · Whitepaper v1.2 · Private deployment
SAOAS
Autonomous Security Decision Orchestration — the decision & response core for security operations
Four-layer architecture Three-tier authorization Dozens of decision modules Human-in-control fallback Full audit trail
AP-WP-SAOAS-2026-001 · Whitepaper v1.2 · Private deployment
TALON deception & attribution ⇄ SAOAS decision orchestration ⇄ response — an automated purple-team loop
Four-Product Suite

AI-native, full-stack purple-team coverage

From external attack surface, to human factors, to AI-system security, to the enterprise command center — each product stands alone; together they reinforce one another.

01ArgusExternal Attack Surface EASM: continuous mapping of exposed assets, ports, certificates and credential leaks into attack-surface snapshots and risk findings. Modular by asset scale · priced after evaluation
02SpectreHuman & Social Engineering AI-generated SE assessment, awareness training and an NPS loop — covering physical SE and deepfake scenarios. Modular by headcount · priced after evaluation
03MirageAI System Security LLM red-teaming, agent behavior monitoring and compliance assessment (aligned to the EU AI Act and other frameworks). Modular by AI-system count · priced after evaluation
04AegisSecurity Operations Center XDR + AI-native SOC: a full capability stack from endpoint, network and intelligence to automated response — governing the suite. Private deployment · scoped to environment
External surface → Human factors → AI security → Command center · modular, bespoke — one company, one price
Capabilities

The full purple-team practice, under one roof

Delivered across the product suite and bespoke engagements — offense and defense as one discipline.

Threat Hunting Incident Response Digital Forensics Threat Intelligence Attacker Attribution Threat Deception Red & Purple Teaming Penetration Testing Social Engineering Assessment LLM Red-Teaming Attack Surface Management Security Automation (SOAR) Detection Engineering Vulnerability Assessment Compliance & AI Governance
Research → Product

Products are not inspiration — they are research, converged

Every product line traces to a specific research program and its core mechanism. Research first, product second — that is what sets AutoPurple apart.

Adversarial security research (AIMP / AIACW)→Attack-surface universality→SAOAS · Mirage
Adversarial multilateration→Tiered source convergence, with confidence→TALON
Live honeynet research→Passive attacker profiling→TALON · Threat Intel
Human factors & social engineering (H-Series)→Persona & scenario assessment→Spectre
Detection & defense engineering→Multi-source detection & response→Aegis · Argus
Research

Before the company, there was the research

"Truth first, evidence-driven — capability is stated to the standard of actual deployment and acceptance, never beyond the evidence."

Martin K. — Independent security researcher / Technical lead, AutoPurple. Research spans AI system security, adversarial machine learning, attacker attribution and human-factor security, engineered into the core of AutoPurple's products.

Adversarial ML AI System Security Attacker Attribution Human Factors Active Defense

AIMP / AIACW Adversarial security research

Mechanisms of LLMs-as-attack-tools and alignment failure, reframed as mechanism-specific alignment. Multiple papers under review at IEEE S&P, USENIX Security, NDSS, ACSAC and SaTML; preprints archived on Zenodo with immediate DOIs.

SAOAS Arch. Theory of orchestration

Axiomatizing attack-surface universality — that all software, security appliances included, is attack surface — and inversion reasoning, the methodological source of the SAOAS decision core.

GEO Adversarial multilateration

Constraint-based network geolocation, plus adversarial reinforcement and attribution through proxies: convergence from region to city network segment to facility, annotated with confidence and physical bounds.

H-Series Human factors & SE

Empirical work on security behavior, persuasion and conviction limits — the methodological basis for Spectre's persona profiling and scenario-based assessment.

Company

Legal, controlled, auditable — throughout

From an independent body of adversarial security research, to an engineered product suite, to flagship whitepapers and private-deployment engagements.

Research phaseAdversarial research programs — AIMP / AIACW / H-Series / multilateration.
H1 2026Product suite engineered — Argus / Spectre / Mirage / Aegis complete.
Jul 2026Flagship whitepapers v1.2 — SAOAS and TALON.
NowPrivate deployment & partner engagements.
Legal Entity
AutoPurple Security Technology Corp.
太原奥紫安全科技有限公司
Approach
Legal · controlled · auditable. Active defense and security automation, governed by "control before autonomy".
Intellectual Property
Core methods protected as trade secrets — not traded for time-limited patent disclosure; software registered under copyright.
Deployment
Primarily private deployment, modular and composable; implementation, customization and ongoing operations support.
Compliance
All attribution derives from adversary interaction within the defender's own deception environment; active measurement is standard network ranging, distinct from intrusion.
Clientele
Large platforms and critical sectors with real needs in adversary profiling and source attribution, and security-collaboration scenarios.
Official Document

Products & engagement — reference

A concise reference to the main questions about our products and the standard for submitting an evaluation request.

AP-DOC-PUB-2026-001 Public · Rev. 1.2
I · About our products
What is AutoPurple?
A purple-team security company: the offensive team's insight (red) and the defensive team's discipline (blue), unified as controlled autonomy. All capability originates from a systematic body of adversarial security research, engineered into products.
What are the flagship platforms?
TALON — threat deception & attribution, the intelligence source and attribution hub of active defense. SAOAS — autonomous security decision orchestration, the controlled decision & response core for security operations. Together they form a closed loop.
What does the four-product suite cover?
Argus external attack surface · Spectre human & social engineering · Mirage AI system security · Aegis enterprise SOC (XDR). Each stands alone; together they reinforce one another.
Is it legal and compliant?
Yes. All deception and attribution occur within the defender's own environment; nothing reaches out of your network boundary. Active measurement is standard network ranging, distinct from intrusion or out-of-bounds counter-attack. Every sensitive action is under tiered authorization and full audit.
How is it deployed and priced?
Primarily private deployment, modular and composable. AutoPurple keeps no standard price list: an independent red-team analysis comes first, then pricing is set from real needs and posture — one company, one price.
II · Ticket submission standard
Before you submit
Select the correct request type (Evaluation, Assistance, Partnership) and your primary product / focus. Provide an accurate company name and a monitored contact email. Describe your current posture honestly — it is the objective basis for scoping.
What happens after you submit
1) You receive an automated confirmation with a ticket reference. 2) Our Chief Information Security Officer reviews your request and sends a non-automated, specific response to your email. 3) A security engineer runs an independent red-team analysis. 4) Scope & posture verification. 5) Bespoke pricing, settled by formal contract.
Confidentiality
All submissions are treated as confidential. Please do not include secrets, credentials, or live vulnerability detail in the form; those are handled under a signed agreement during engagement.
Download the full Product Technical Overview ↓ PDF AP-DOC-PUB-2026-002 · Public · No confidential detail
产品与市场说明（中文 · 商业版） ↓ PDF AP-DOC-PUB-2026-003 · Public · Chinese · No confidential detail
Engagement

One company, one dedicated engineer — high-touch bespoke security

AutoPurple keeps no standard price list. Each client is paired with a dedicated security engineer; an independent red-team analysis comes first, then pricing is set from real needs and posture. Modular architecture is precisely why every engagement differs.

01Red-Team AnalysisA consulting engineer runs an independent security analysis of your environment — seeing the real attack surface.
02Scoping & NegotiationFrom the findings and your business reality, we define the required modules and service boundaries together.
03Posture VerificationThe real security posture and adversary intensity are verified as the objective basis for the plan.
04Bespoke PricingPriced to the actual modular scope — one company, one price, settled by formal contract.
Let an engineer see your real attack surface first
An independent evaluation by our engineer and officer comes before any plan or pricing. This is the only way to engage AutoPurple.
Engineer reply within 48h Evaluation first, plan after Confidential throughout
Submit an evaluation request
Tell us about your environment. Every submission is reviewed by our engineer, who replies within 48 hours. Fields marked * are required.
or email [email protected]