Comprehensive Feature Set

Governed Tools for
Clinical-Legal Review

From chronology drafts to source-linked report work product, explore the Case Workflow that helps Legal Nurse Consultants organize evidence, review candidate findings, and preserve professional judgment.

AI Chronology Builder

Turn supported medical-record files into chronology drafts that can be reviewed and refined by your team.

Supported file ingestion
Import supported PDFs, images, and text files for extraction and review.
Auto-Dating & Sorting
Extract dates and organize a reviewable clinical chronology
Entity Extraction
Extract supported clinical entities with visible confidence signals

Evidence Analysis

AI-assisted analysis of medical evidence to surface candidate insights and potential issues for professional review.

Gap review
Surface possible missing care periods and documentation gaps for review.
Medication Tracking
Track medication changes, interactions, and adverse events
Procedure Timeline
Compare documented care with applicable Standards of Care for review

CRCM Pattern Matching

Comprehensive Risk and Compliance Model logic engine that aligns clinical patterns with legal causes of action.

Pattern Library
Pre-built patterns for common causes of action
Cause-of-action review
Map candidate clinical findings to legal theories for qualified review.
Confidence Scoring
Confidence signals that help route pattern matches for review; not accuracy guarantees

ForenCoT Causal Review

Structured, reviewable causal analysis for forensic workflows. BluQube presents source context and reasoning signals; qualified professionals make the judgment.

Reviewable outputs
Inspect the source context and reasoning signals associated with candidate events or patterns.
Audit and provenance records
Record supported user and analysis events for later review.
Draft explanations
Prepare reviewable language for professional editing and report assembly.

GraphRAG Pattern Analysis

Build co-occurrence graphs from CRCM pattern matches to discover hidden clinical relationships and community clusters. Graphs are built per case today; portfolio-wide co-occurrence graphing is In Development.

Pattern Graphs
Visualize how clinical risk factors cluster together within a case
Community Detection
Automatically identify groups of related patterns for focused review
Interactive Visualization
Risk-factor heatmap and per-case pattern co-occurrence graph in the dashboard

Entity Extraction & Recipe Search

Schema-based extraction identifies diagnoses, anatomical sites, and medications from clinical text. Multi-strategy recipe search combines entity expansion, community context, and semantic similarity.

Medical Entity Extraction
Extract and normalize clinical entities with confidence scores
Recipe-Based Search
Predefined search strategies with reciprocal rank fusion for richer results
Entity Resolution
Deduplicate and normalize entities across documents for consistent analysis

Expert Report Generation

Generate first drafts of expert reports based on analyzed chronologies.

Template-Based
Customizable templates for different case types and jurisdictions
Citation Validation
Re-check cited medical records and evidence sources against the case record
Professional formatting
Use structured headings and supported export formats as a starting point for professional editing.

Collaborative Workspace

Role-based collaboration for legal teams, Legal Nurse Consultants, and qualified reviewers.

Role-Based Access
Granular permissions for attorneys, LNCs, and medical experts
Activity History
Review supported user actions and document events captured by the application.
Audit logs
Review supported user actions and document events captured by the application.

Full explanations

Product Highlights in Depth

Every highlight explained the way a reviewer should expect: what it does, how it works, why it matters, and what sets it apart.

Evidence Integrity Chain

What it does: Every registered piece of evidence gets a cryptographic fingerprint and a provenance record, and the platform re-checks both before trusting the file.
How it works: At registration the platform computes a SHA-256 hash of the file and records provenance events — who added what, and when. Before analysis or export, the stored hash is recomputed and compared, so a file that changed after registration fails its integrity check instead of silently passing through.
Why it matters: In litigation someone will ask whether a document is the same one that was received. A verifiable hash and a provenance record let the reviewer answer with evidence rather than memory.
What differentiates: Generic document stores track filenames and timestamps. Here integrity is re-verified at use time, and a failed check blocks trust instead of going unnoticed.

Citation Validation

What it does: The platform automatically re-checks that each AI finding's cited sources are present and verifiable in the case record.
How it works: Every AI analysis stores its citation references. A validation pass re-queries each citation against the case- and organization-scoped record set and returns a per-citation result. If the scope cannot be resolved, validation fails closed — it returns nothing rather than guessing.
Why it matters: A finding whose citation does not hold up is a liability at deposition. Automated citation checking means the reviewer spends time on professional judgment, not on hunting page cites.
What differentiates: Citation formatting is table stakes. Automated re-verification of citations against the record is the difference between a finding that looks cited and one that is traceable.

Send-Gate Screening

What it does: No client deliverable leaves the platform without a named human approval — enforced structurally, not by policy.
How it works: Producers enqueue deliverables; a reviewer screens and approves them in an inbox. A single delivery worker transmits, and its storage query only returns approved rows — unscreened items are never returned at all, so no downstream bug can send them. Before transmission the integrity seal is re-verified, and an item mutated after approval dead-letters instead of sending.
Why it matters: Anything that reaches a client or an opposing party carries the firm's responsibility. The gate guarantees every transmitted item has a human decision and an intact integrity seal behind it.
What differentiates: Most platforms rely on workflow conventions that can be bypassed. Here the storage layer itself makes an unscreened send impossible.

CRCM Pattern Analysis

What it does: Maps clinical findings to candidate legal theories using a pattern library built for medical-malpractice review.
How it works: A rules engine matches extracted clinical factors against CRCM patterns. A risk-factor heatmap shows the overall assessment, the worst single factor, and the matched patterns for the case, and a per-case co-occurrence graph shows how risk factors cluster together. Portfolio-wide graphing is In Development.
Why it matters: It turns "something looks off" into a named, reviewable pattern hypothesis the consultant can validate against the record — candidate findings, not conclusions.
What differentiates: Generic legal AI summarizes records. Here the analysis follows a named clinical-compliance model, and every match carries the review status a professional workflow needs.

ForenCoT Reasoning & Causal Review

What it does: A structured workflow that assembles reviewable analysis chains, including causal review, instead of presenting free-form answers as conclusions.
How it works: Chain building, selection, assembly, and quality checks run as separate stages, and causal review chains are persisted per case. The reviewer-facing page shows the source context and reasoning signals attached to each candidate event or pattern.
Why it matters: An expert opinion must be explainable step by step. A chain a reviewer can inspect, challenge, and correct is defensible in a way a black-box answer is not.
What differentiates: Many products hide their reasoning. Here the reasoning chain is a first-class, persisted artifact the reviewer can audit before anything becomes work product.

AI Chronology Builder

What it does: Turns supported medical-record files into a dated, reviewable chronology draft your team refines — not a finished document it has to distrust.
How it works: Extraction, entity and event detection, timeline building, and narrative generation run as one integrated pipeline over extracted record text that keeps page-marker conventions, so drafted events stay traceable to the pages they came from. The case timeline is served per case for review and refinement.
Why it matters: The chronology is the spine of every medical-legal review. A machine-drafted, human-corrected timeline beats starting from a blank page — and beats reading 2,000 pages twice.
What differentiates: Because the draft is grounded in page-level source markers, each event can be checked against its origin instead of taken on faith.

Schema-Based Entity Extraction

What it does: Pulls diagnoses, anatomical sites, medications, and clinical events out of unstructured nursing notes, each with a confidence score.
How it works: A schema-based extractor emits typed entities with confidence scores between 0 and 1, and medical schema normalization resolves the same real-world concept to the same entity across documents, so deduplication and cross-document analysis work from one consistent view.
Why it matters: Everything downstream — chronology, pattern matching, terminology review — depends on consistent entities. Inconsistent extraction corrupts analysis silently; confidence scores make the uncertainty visible instead.
What differentiates: Confidence is used as a routing signal for review, not an accuracy guarantee — the same honest posture the rest of the product takes.

Recipe-Based Multi-Strategy Search

What it does: Predefined search recipes combine four retrieval methods into one ranked result set, so a case question gets one good answer list instead of four partial ones.
How it works: Each recipe dispatches entity expansion, community context, semantic similarity, and pattern matching against real provider implementations, then merges the ranked lists with reciprocal rank fusion — a principled blend rather than an arbitrary weighting.
Why it matters: Medical-record search fails when the chart's vocabulary differs from the lawyer's question. Fusing independent strategies reduces the chance a critical page hides behind a single vocabulary.
What differentiates: Single keyword search is the baseline. Four independent strategies fused with a documented rank-fusion method is a different retrieval posture — and the recipes are predefined, so reviewers get consistent searches.

Intake Quality & Reviewer Disposition

What it does: Detects record-quality problems at intake, and every finding goes through a reviewer disposition workflow before it is trusted or ignored.
How it works: Detectors quarantine noise pages (MAR calendar grids, table-of-contents pages, bare code dumps), flag duplicate pages and encounters, and detect mixed-patient pages. In the dashboard, a reviewer accepts or flags each finding, and the decision is recorded with actor and note.
Why it matters: Garbage in, garbage out is the real failure mode of record analysis. Quarantining noise before analysis protects every number and quote that comes after it.
What differentiates: Many pipelines extract blindly. Here intake quality is a first-class stage with an auditable human decision trail — who accepted what, and why.

Terminology Review

What it does: Shows the case's standardized clinical terminology — LOINC document classifications and diagnosis assertions — with polarity, review status, and correction lineage.
How it works: A read-only route surface serves the terminology store: it exposes GET and nothing else, so the review desk can see structured clinical language with no risk of mutating it from that screen. Corrections carry lineage, so a changed assertion shows what it was corrected from.
Why it matters: Normalized clinical language is what makes records searchable and comparable across providers. Reviewers need to see what was extracted — including what was corrected — before relying on it.
What differentiates: The display-only construction is the point: this surface cannot silently alter clinical data, by design rather than by permission.

Care-Cost Quantification & Chart

What it does: Damages sections can quantify future care costs from structured care-item inputs, with a cumulative cost-of-care chart embedded in the report — when those inputs are on record for the case.
How it works: The quantification core does all money arithmetic in exact decimal arithmetic, each line quantized once to the cent — no floating-point drift between calculation and report. Ingestion is fail-closed at the row level, and the report embeds a deterministic, server-rendered cumulative curve, so the same inputs always produce the same chart.
Why it matters: Cost of care is the financial heart of serious-injury damages. Exact arithmetic means the numbers in the report can be reproduced line by line if they are challenged.
What differentiates: Spreadsheet summaries drift and are hard to audit after the fact. This core is deterministic and reproducible from the inputs it names — the chart is generated from the same math as the table.

Expert Report Studio & D-Series Exhibits

What it does: Drafts expert reports from the analyzed record and serves the D-series exhibit files already prepared for the case.
How it works: The report studio exports to HTML, PDF, and DOCX through an asynchronous export queue, so long renders never block the review. The exhibit service lists and serves D-series files per case behind authentication and a filename gate. Report state includes a human attestation gate — a deliverable is marked ready only when a person has attested to it with recorded scope.
Why it matters: The deliverable is the product. Drafts that arrive in editable, exportable, citable form compress the distance between analysis and billable work product.
What differentiates: Report generation is common. A human sign-off gate built into the report's own state — rather than an email thread — is not.

AI Governance: Circuit Breaker & Assurance Tiers

What it does: A control system that watches AI output quality in production and can pause the AI pipeline automatically — governance as a product feature, not a slide.
How it works: A circuit breaker monitors accuracy over a sliding window of recent decisions. If accuracy drops below a configurable threshold the circuit opens and auto-pauses the system — every new AI call is refused while paused — and after a recovery timeout it tests recovery before resuming. In parallel, a risk-tier engine classifies each AI action and routes it to a defined handling action, and blocking validation exceptions raise the routing tier rather than passing silently.
Why it matters: In clinical-legal work, an AI system that degrades quietly is more dangerous than one that is not there. Automatic pause on degradation bounds the blast radius of a bad model day.
What differentiates: Most products publish quality claims. This one ships the enforcement mechanism that refuses every new AI call when it underperforms — and records the pause.

Governed Process: SOP Pipeline & Disclosure Register

What it does: Case work follows a 17-phase standard-operating-procedure pipeline, and customer-directed disclosures are recorded in a HIPAA §164.528 disclosure-accounting register.
How it works: The phase manifest declares each phase's dependencies; the pipeline orders phases topologically and gates uploads before any phase runs. The disclosure register is append-only — every disclosure is a new row recording recipient, what was disclosed, when, and under what authority. The register records disclosures; the decision to disclose always remains the organization's.
Why it matters: Defensible work product starts with a defensible process. When a client or regulator asks who received what and when, the answer is a query, not a reconstruction.
What differentiates: Checklists describe process. Here the process order is enforced in code, and the disclosure record is durable, queryable, and tamper-evident by construction.

BluQube vs Manual Process

Use the workflow to reduce repetitive work; actual results depend on record quality, configuration, and review.

Feature	Manual Process	BluQube
Chronology Creation	Manual review	AI-assisted draft
Evidence Analysis	Manual review	Candidate findings
Pattern Identification	Manual review	AI-assisted
Report Drafting	Manual drafting	AI-assisted draft
Quality control	Professional review	Required
Case cost	Varies	Depends on plan and usage

See the Case Workflow in Practice

Explore how BluQube organizes evidence, analysis drafts, professional review, and source-linked work product.

Governed Tools for Clinical-Legal Review

Feature details

AI Chronology Builder

Evidence Analysis

CRCM Pattern Matching

ForenCoT Causal Review

GraphRAG Pattern Analysis

Entity Extraction & Recipe Search

Expert Report Generation

Collaborative Workspace

Product Highlights in Depth

Evidence Integrity Chain

Citation Validation

Send-Gate Screening

CRCM Pattern Analysis

ForenCoT Reasoning & Causal Review

AI Chronology Builder

Schema-Based Entity Extraction

Recipe-Based Multi-Strategy Search

Intake Quality & Reviewer Disposition

Terminology Review

Care-Cost Quantification & Chart

Expert Report Studio & D-Series Exhibits

AI Governance: Circuit Breaker & Assurance Tiers

Governed Process: SOP Pipeline & Disclosure Register

BluQube vs Manual Process

See the Case Workflow in Practice

Governed Tools for
Clinical-Legal Review