See the Pattern. Know the Truth.

AI-powered behavioral analysis that maps institutional communications against clinically-grounded patterns. Peer-reviewed methodology. Mathematical transparency. Your text never lies.

or
Level 2 Coherence Analysis · Enterprise reports available

What You'll See

A real example of behavioral scoring on institutional language.

Input Text
"Due to organizational restructuring, your position has been eliminated effective immediately. This decision is final and not subject to appeal."
Unilateral Exercise of Power87%
Retaliatory Targeting72%
Bureaucratic Stonewalling68%
Legitimate Administration23%
Healthy Self-Advocacy12%
ACTOR: Institution → RECIPIENT: Employee ⚠ Adverse action 4 days after formal complaint · TPS: 0.887

How It Works

Three steps from raw text to mathematical clarity.

📋

Paste

Copy any email, letter, HR notice, legal document, or text message into the analyzer.

🧬

Map

Your text is embedded into 768-dimensional space and measured against clinically-grounded behavioral anchors.

📊

Know

See mathematical proximity scores showing exactly how your text aligns with established patterns of institutional behavior.

Peer-Reviewed Methodology

Validated by research in Nature Human Behaviour, NeurIPS, JMIR, and NIH.

Nature Human Behaviour

Semantic Embeddings for Psychological Measurement

Wulff & Mata, 2025

"Embeddings preserve construct relationships in vector space"

View Paper →
NeurIPS

Personality in LLM Latent Space

Suh et al., 2024

"74.3% variance explained via SVD on trait descriptors"

View Paper →
PsyArXiv

Gaslighting Detection via Emotion Vectors

Reilly et al., 2024

"85% accuracy via composite emotion vectors"

View Paper →
JMIR AI

Psychiatric Classification from Embeddings

Shewcraft et al., 2025

"AUC 0.89–0.97 across 7 disorders"

View Paper →
JMIR Med Inform

LLMs vs. Clinicians on Diagnosis

Sun et al., 2026

"71.7% accuracy from 9,923 clinical records"

View Paper →
npj Mental Health

Fairness in Behavioral AI

Wang et al., 2026

"Equity-aware prediction framework"

View Paper →

What We Add

Published science validated the foundation. Here's what we built on top.

🎯

Asymmetric Directionality

Most AI tools ask "Is this abusive?" We ask "Who is doing what to whom?" — because the same words describe very different realities depending on who holds the power.

Triple Extraction

We decompose text into structured actions — Subject → Relation → Object — and score each independently.

⏱️

Temporal Proximity Score

TPS = e−λt. A termination 3 days after a complaint scores 0.914. 90 days later: 0.067. The math quantifies what lawyers know intuitively.

🏛️

Institutional Power Anchors

DSM-5 describes individual pathology. Institutions aren't individuals. We built anchors for stonewalling, unilateral power, and retaliatory targeting.

🔒

Local-First Privacy

Your text never leaves your machine. Full local analysis mode — zero cloud, zero logging, zero third-party access.

Built for Every Arena Where Power Meets Paper

From courtrooms to newsrooms — anywhere institutions put it in writing.

⚖️

Employment Law

Detect retaliation patterns, wrongful termination timelines, and whistleblower suppression.

🏥

Medical Malpractice

Analyze hospital communications around complaint timelines. Detect administrative cover-up patterns.

📰

Investigative Journalism

Pattern analysis across leaked documents. Surface institutional abuse invisible to the human eye.

🏛️

Congressional Oversight

Analyze agency testimony, whistleblower submissions, and institutional correspondence at scale.

🔍

Antitrust

Detect coordinated stonewalling and cartel-like communication patterns across corporate correspondence.

🛡️

Healthcare Fraud

Map bureaucratic obstruction patterns in billing disputes and coverage denials.

Choose Your Level

From a single analysis to enterprise-scale deployment.

Individual
$29
/analysis
  • Full behavioral scoring
  • Directionality analysis
  • Temporal proximity scoring
  • Exportable PDF report
Institutional
Contact
 
  • Congressional testimony analysis
  • Whistleblower pattern detection at scale
  • API access
  • Custom deployment
  • On-premise option

This tool wasn't built in a lab.

It was built by someone who needed it — to analyze real institutional communications, detect real retaliation, and prove real patterns. The methodology emerged from necessity, was refined against real data, and is now validated by peer-reviewed science.

— Brent Porter, Founder