6-Agent Pipeline | Privacy-First | Local AI

Law Lite

AI Due Diligence for Corporate & Securities Law

Automated compliance analysis with privacy-first architecture. Six specialized AI agents process corporate documents, detect risks, and generate executive summaries — with PII anonymized before any AI sees it.

Six Agents, One Due Diligence

Each agent is specialized. Documents flow through ingestion, anonymization, compliance, risk analysis, summary generation, and report assembly — with anonymization boundaries protecting PII at every step.

Upload Lex Petra ANON Gavel Reese Portia DE-ANON Sterling Done
01 Tier 1 Local

Lex

Document Ingestion & Parsing

Ingests corporate documents — merger agreements, securities filings, board resolutions, debt covenants. Classifies and parses each document for downstream analysis.

Cost: $0
02 Tier 1 Local

Petra

PII Detection & Anonymization

Scans all parsed documents for personally identifiable information — names, dates, addresses, SSNs, account numbers. Replaces PII with anonymized tokens before any AI analysis.

Cost: $0
03 Tier 1 Local

Gavel

Compliance Rule Engine

Runs 30 deterministic compliance rules across five practice areas. Checks regulatory requirements, filing deadlines, disclosure obligations, and governance standards.

Cost: $0
04 Tier 2 Local/Cloud

Reese

Risk Analysis & Scoring

Multi-dimensional risk assessment using local LLM. Analyzes cap table structures, covenant compliance, regulatory exposure, and transaction risk factors.

Cost: Low
05 Tier 2 Local/Cloud

Portia

Executive Summary Generation

Generates comprehensive executive summaries with risk ratings, key findings, and actionable recommendations. Works on anonymized data only.

Cost: Low
06 Tier 1 Local

Sterling

Report Assembly & De-anonymization

Assembles the final due diligence report, de-anonymizes all PII tokens back to real values, generates closing checklists, and purges temporary mappings.

Cost: $0

Everything You Need for AI Due Diligence

Built for corporate and securities law professionals who want AI speed without sacrificing compliance, privacy, or control.

🧠

Smart Tools

6 specialized analysis tools — cap table validator, covenant analyzer, risk scorer, red flag detector, checklist generator, and compliance mapper — each purpose-built for corporate law workflows.

📊

Cap Table Analysis

Validates ownership structures, identifies dilution risks, checks voting rights consistency, and flags discrepancies in shareholder agreements and equity plans.

📜

Covenant Analyzer

Parses debt covenants and loan agreements. Checks financial ratio compliance, identifies cross-default triggers, and monitors covenant breach thresholds.

Risk Scoring

Multi-dimensional risk assessment across legal, financial, regulatory, and operational dimensions. Weighted scoring with configurable thresholds and severity levels.

30 Compliance Rules

Comprehensive regulatory checks covering SEC filing requirements, corporate governance standards, disclosure obligations, antitrust thresholds, and securities regulations.

🎯

5 Practice Areas

Purpose-built rule sets for M&A, Securities, Corporate Governance, Debt Finance, and Regulatory — each with specialized compliance checks and risk factors.

🔒

PII Anonymization

Zero-knowledge document processing. All personally identifiable information is detected and replaced with tokens before AI analysis. AI never sees real names, dates, or identifying data.

💻

Ollama Local-First

Runs fully offline with Ollama local models. Sensitive documents never leave your premises. Cloud AI is an optional fallback, and only processes anonymized content.

🚩

Red Flag Reports

Automatic identification of legal red flags — undisclosed liabilities, regulatory violations, governance gaps, missing filings, and material risk factors requiring immediate attention.

Closing Checklists

AI-generated transaction readiness checklists tracking every required document, approval, filing, and condition precedent for deal closing.

📂

Demo Cases

5 built-in practice scenarios covering M&A due diligence, securities offering review, governance audit, debt finance analysis, and regulatory compliance check.

📋

Checklist Generator

AI-powered task generation creates detailed checklists from document analysis. Tracks completion status, assigns priorities, and identifies blocking dependencies.

AI Never Sees Real Names, Dates, or Identifying Information

The anonymization boundary between Petra and Gavel ensures that real personally identifiable information is never processed by AI analysis agents. Only anonymized tokens are analyzed.

On-Premises (Document Intake)
Lex Petra
🔒 Anonymize John Smith → PARTY_A
2024-03-15 → DATE_A
AI Analysis (Anonymized Only)
Gavel Reese Portia
🔓 De-anonymize PARTY_A → John Smith
Mapping purged
On-Premises (Delivery)
Sterling

Ephemeral Mappings

Anonymization mappings exist only in memory during processing. They are never written to disk and are purged immediately after report delivery.

Local-First Processing

Lex and Petra run entirely on your hardware with full PII access. Gavel, Reese, and Portia analyze only anonymized data. Sterling de-anonymizes locally.

Audit Trail

Every step is logged with timestamps. The audit records that anonymization was applied, without storing the actual mapping between real and anonymized values.

See the Pipeline in Action

Frequently Asked Questions

What is Law Lite?

Law Lite is an AI-powered due diligence platform for corporate and securities law. It uses six specialized AI agents to process legal documents, check compliance, assess risk, and generate executive summaries — all with privacy-first PII anonymization.

What practice areas does it cover?

Law Lite covers five practice areas: M&A (mergers and acquisitions due diligence), Securities (offering reviews and SEC compliance), Corporate Governance (board resolutions and governance audits), Debt Finance (covenant analysis and loan agreement review), and Regulatory (compliance checks and filing requirements).

How does the pipeline work?

Six agents process documents sequentially: Lex ingests and parses documents, Petra detects and anonymizes PII, Gavel runs 30 compliance rules, Reese performs risk analysis and scoring, Portia generates executive summaries, and Sterling assembles the final report and de-anonymizes PII back to real values.

How is privacy handled?

Petra detects all personally identifiable information (names, dates, addresses, account numbers) and replaces them with anonymized tokens before any AI analysis occurs. Gavel, Reese, and Portia only ever see anonymized data. Sterling de-anonymizes the final report locally, and all mappings are purged immediately after delivery.

What AI models does it use?

Law Lite uses Ollama local models (Qwen, Nemotron) as the primary AI engine — all processing happens on your hardware. Claude cloud AI is available as an optional fallback for complex analysis, but only ever receives anonymized data with no real PII.

Can it work offline?

Yes. Law Lite is fully functional with Ollama local models running on your own hardware. No internet connection is required. The entire pipeline — from document ingestion through report delivery — runs entirely on-premises in air-gapped mode.

What are Smart Tools?

Smart Tools are six specialized analyzers built into the pipeline: Cap Table Validator (ownership structure checks), Covenant Analyzer (debt covenant compliance), Risk Scorer (multi-dimensional risk assessment), Red Flag Detector (automatic issue identification), Checklist Generator (AI-powered task generation), and Compliance Mapper (regulatory requirement tracking).

How do I deploy Law Lite?

Law Lite supports two deployment modes: Helm chart for Kubernetes clusters (recommended for enterprise), and Docker for standalone deployment on a single machine. Both modes include Ollama local model support and can be configured for fully air-gapped operation.

See Law Lite in Action

Book a demo to see the full pipeline process corporate legal documents with privacy-first AI analysis.

Ajay Tyagi, VP Sales — ajay@dkube.io

US Dkube Contact: +(1) 408 430 2503