I help organizations understand, audit, and navigate complex systems at the intersection of artificial intelligence, law, and public policy.
End-to-end research, analysis, and advisory across the intersections of AI, law, data, and public accountability.
Evaluate LLM reliability, bias, and failure modes. Design duty-of-care frameworks and conduct comparative model analysis across deployment contexts.
Analyze legislation, court cases, and regulatory systems. Translate legal complexity into actionable insights for technical and policy audiences alike.
Build reproducible pipelines for large-scale document analysis — entity extraction, sentiment, co-occurrence networks, and targeting pattern detection.
Audit misinformation dynamics, trace narrative propagation, and design evidence-based communication strategies grounded in empirical methodology.
I audit complex documents to reveal what they are actually doing beneath the surface.
Not summaries. Not opinions. Structural, data-backed analysis of language, intent, and internal consistency.
Using AI as an analytical instrument—not a content generator—I surface the signals that traditional review misses.
Internal inconsistencies that undermine stated positions or expose unstated assumptions.
Conflicting worldviews coexisting within a single document — often unacknowledged.
Measurable signals of urgency, threat, and authority used to steer interpretation.
Where tone, framing, or intent changes across sections — and what that signals.
Patterns of language that suggest coordination, fragmentation, or deliberate manipulation.
Most high-impact documents today are too large to critically read in full, present themselves as internally consistent, and rely on rhetoric that goes unmeasured and unchallenged. Traditional analysis misses this.
I make the invisible structure visible — and defensible.
Define analytical dimensions tailored to the document — power, legitimacy, urgency, sentiment, and more.
Process entire documents at scale — thousands of text segments — without sacrificing precision.
Cross-check outputs across systems to reduce bias and error. No single-model dependency.
Identify topic clusters, narrative shifts, and internal fractures across the document body.
Deliver clear, evidence-backed findings — not raw data. Structured for both technical and non-technical stakeholders.
Identify internal contradictions and ideological splits in regulatory and legislative texts.
Detect narrative framing and strategic messaging in advocacy and policy documents.
Surface inconsistencies in disclosures, filings, or strategic positioning documents.
Analyze tone, framing, and structural bias in academic and institutional work.
Most AI work today generates content. This does the opposite. Every result is human-directed, cross-validated, and evidence-based — with no automation without oversight, no black-box conclusions, and no single-model dependency.
You are not buying AI output.
You are hiring a Principal Investigator using AI as a high-resolution analytical instrument.
Deep analysis of one report, policy, or dataset. Full pipeline applied to a single high-stakes document.
Side-by-side structural audit of multiple documents — competing policies, organizations, or time periods.
Continuous analysis for organizations tracking narrative risk, policy shifts, or evolving regulatory landscapes.
Clarity is leverage. Structure is truth.
In an environment flooded with generated content and unverified claims, I provide both.
Every conclusion is traceable and defensible. Pipelines are built for reproducibility from the start, not retrofitted.
No employer, no funder, no conflicts of interest. Analysis goes where the evidence leads.
Code, data, and reasoning are structured so others can replicate, audit, and challenge the work.
Complex systems become actionable insight. Reports are written for both technical and non-technical stakeholders.
Research infrastructure built for rigor, reproducibility, and scale.
Available for document audits, research projects, and advisory engagements.
contact@narrovue.com