Skip to content
AL | Apatheia Labs
Forensic Intelligence

Expose institutional dysfunction through evidence.

Apatheia Labs builds tools for institutional accountability analysis. Systematic methodology. Full audit trails. No hallucination.

Our flagship platform, Phronesis, detects contradictions, constructs timelines, and maps bias patterns across complex document bundles — entirely on your machine.

Join WaitlistHow It Works
Analysis Findings
3 critical issues detected

Temporal Contradiction Detected

Report claims assessment occurred on 15 March, but referenced document dated 22 March.

TEMPORALDoc 14 → Doc 23

Selective Omission Identified

Expert report excludes 3 paragraphs from source that contradict conclusions.

SELECTIVEBias: +1.0

The Problem

Why Institutions Fail Without Accountability

Institutional dysfunction persists because the evidence is fragmented, the patterns are hidden, and manual analysis cannot scale to document volumes that matter.

Claims Propagate Without Verification

A single unverified statement can be adopted by multiple agencies, accumulating apparent authority through repetition rather than evidence. By the time it reaches a decision-maker, the original uncertainty has been laundered into certainty.

Contradictions Hide in Volume

When case files span hundreds of documents, internal contradictions become invisible. A date mismatch on page 47 that undermines a conclusion on page 312 will never be caught by human review alone.

Selective Citation Goes Undetected

Reports that quote sources selectively—omitting context that contradicts their conclusions—appear authoritative. Without systematic source comparison, cherry-picking masquerades as thorough analysis.

Methodology

Systematic Adversarial Methodology

S.A.M. is the framework for reading institutional documents “against the grain”—treating every claim as requiring verification rather than acceptance.

Most document review assumes good faith: that claims are accurate, that sources are cited fairly, that timelines are consistent. S.A.M. inverts this assumption. It systematically tests every assertion, traces every claim to its origin, and maps how information flows between institutions.

1ANCHOR

ANCHOR is the foundational phase of the S.A.

2INHERIT

INHERIT tracks how claims propagate across institutional boundaries without independent verification. When one agency’s assessment is adopted by another without re-evaluation, errors compound and gain apparent authority.

3COMPOUND

COMPOUND documents how claims gain weight through repetition rather than verification—what is termed "authority laundering." A statement repeated by four professionals appears more credible than one stated by a single source, even when all four trace to the same unverified origin.

4ARRIVE

ARRIVE maps catastrophic outcomes back to their originating false premises, establishing but-for causation: would the outcome have occurred if the anchor claim had been verified? This phase connects the analytical work of phases 1–3 to real-world consequences, demonstrating how initial errors cascaded into final decisions..

CASCADE: Eight Contradiction Types

A taxonomy for classifying inconsistencies detected across document corpora

Built for Trust

Evidence-Grade Analysis

Systematic methodology with full audit trails. Every finding traceable to source.

Trust Architecture

Built-in safeguards against AI hallucination and unsupported claims

  • Source Citation Required

    Every claim traced to [Doc:Page:Para] references

  • Confidence Scoring

    FACT vs INFERENCE vs SPECULATION distinction

  • Verification Layer

    NLI-based claim validation

  • No Hallucination Risk

    Rejects unsupported assertions

Proven Capabilities

Tested on real investigations with 1,000+ page document bundles

  • Directional Bias Scoring

    +1.0 to -1.0 scale measuring systematic imbalance

  • Framing Ratio Detection

    Statistical significance testing with p-value calculations

  • Timeline Reconstruction

    Multi-year analysis across 1,000+ page bundles

  • Omission Pattern Analysis

    Gap detection with bias direction calculation

Capabilities

Analysis Engines

Six specialized engines work in concert to surface patterns that human review would miss. Each engine can run independently or as part of a comprehensive analysis pipeline.

Entity Resolution

Creates a unified identity registry across your document corpus. Resolves aliases, tracks role changes, and maps relationships between individuals.

Alias DetectionRole TrackingRelationship Mapping

Temporal Parser

Constructs unified timelines from scattered date references. Identifies gaps, overlaps, and impossible sequences that indicate error or manipulation.

Date ExtractionTimeline ConstructionGap Detection

Argumentation

Deconstructs claims into their logical components using the Toulmin model. Identifies unsupported assertions, missing warrants, and logical gaps.

Claim ExtractionWarrant AnalysisGap Detection

Contradiction

Systematically detects inconsistencies across your document corpus using the CASCADE 8-type framework. Surfaces conflicts that undermine institutional conclusions.

CASCADE DetectionSource ComparisonConflict Mapping

Bias Detection

Performs statistical analysis of directional bias in reporting. Calculates omission ratios, measures framing imbalance, and quantifies selective presentation.

Directional ScoringOmission AnalysisFraming Metrics

Accountability

Maps statutory duty violations to specific actors and regulatory bodies. Links evidence to relevant professional standards and complaint pathways.

Duty MappingActor AttributionRegulatory Routing

Capabilities

What Phronesis Delivers

Professional-grade forensic analysis, running entirely on your machine.

Included in Phronesis

  • Contradiction Detection

    Find logical inconsistencies across documents

  • Timeline Construction

    Automatically build chronologies from dates

  • Entity Mapping

    Track people, organizations, and relationships

  • Professional Exports

    Generate court-ready reports and summaries

  • Bias Detection

    Identify selective citation and framing

  • Argumentation Analysis

    Map claim structures and evidence chains

  • Regulatory Pathways

    Pre-configured complaint generation for Ofcom, ICO, LGO, HCPC, GMC, SRA

Desktop application. Local-first. Your documents never leave your machine.

Research Hub

Open Methods, Public Standards

47 articles across 13 categories. Phronesis is grounded in professional investigation frameworks, evidence standards, and quality control methods. The research hub is where the sources, comparisons, and reasoning behind every engine are published.

Methodologies

14

Full methodology briefs and quickstart guides for each professional investigative framework.

Foundations

6

Core theoretical frameworks underpinning Phronesis — S.A.M. methodology, CASCADE theory, contradiction taxonomy, epistemology, and empirical validation studies.

Data Science

3

Research on statistical validation, causal inference, and data quality practices for evidence review.

Ethics

3

Research on professional ethics, research ethics, and medical ethics for institutional accountability.

Interdisciplinary

3

Cross-domain syntheses integrating multiple research areas for comprehensive analysis.

Legal

3

Research on evidence law, administrative law, and legal standards for institutional accountability.

Browse All 47 Articles

The Future

What I’m Building

Phronesis is just the beginning. A suite of forensic intelligence tools designed to make institutional accountability accessible to everyone.

Concept

Aletheia

Political Analytics Platform

Apply the same forensic rigor to political discourse. Analyze claims from news outlets, YouTubers, influencers, and political figures. Track accuracy over time and surface contradictions.

YouTube AnalysisTwitter/X MonitoringPodcast SupportClaim DatabaseHistorical Accuracy
Research

Nous

Adaptive Analytics Platform

Describe what you want to understand. The system builds the right analytical approach for your goal. Structure emerges from purpose, not templates. No databases to configure, no schemas to learn.

Intent CaptureDynamic PipelinesSession MemoryExploratory ModeZero Setup
In Development

Obsidian Integration

Deep integration with Obsidian for researchers who live in their knowledge bases. Sync findings directly to your vault with bidirectional linking.

Vault SyncWikilinksYAML Frontmatter
In Development

MCP Server

Model Context Protocol server enabling Claude and other AI assistants to directly query your analysis database and run forensic operations.

Claude IntegrationTool DefinitionsLocal-First
Next Up

Advanced Omission Detection

Compare any document against its source materials to identify what was excluded. Calculate directional bias scores and map selective citation patterns.

Source ComparisonBias ScoringGap Mapping
Next Up

Expert Witness Analyzer

Evaluate expert reports against professional standards. Check FJC compliance, assess scope boundaries, and identify opinions that exceed expertise.

FJC ComplianceScope AnalysisStandard Mapping
Planned

Documentary Comparison

Compare broadcast or published content against source materials. Quantify what was included versus omitted and measure editorial bias in framing.

Broadcast AnalysisFraming MetricsSource Matching
Planned

Cross-Institutional Mapping

Detect hidden coordination between supposedly independent agencies. Identify shared language, circular citations, and pre-disclosure information flow.

Coordination DetectionCitation NetworksPattern Matching

Coming Soon

Join the Waitlist

Be the first to know when Phronesis launches. Get early access and exclusive updates.

Get Notified at Launch

Professional-grade forensic document analysis, running entirely on your machine.

No spam. Unsubscribe anytime.

What You'll Get

  • ✓Early access to Phronesis before public launch
  • ✓Exclusive updates on development progress
  • ✓Priority support and feedback opportunities
  • ✓Local-first: Your documents never leave your machine