Features
Everything you need to process, understand, and analyze legal documents — powered by AI, running locally.
14 AI-Powered Analysis Tools
Every tool runs locally through the Model Context Protocol. Connect your preferred AI assistant and start analyzing.
Contradiction Detection
Find inconsistencies between filings, depositions, and declarations.
Timeline Extraction
Automatically build chronologies from dates mentioned across documents.
Entity Recognition
Identify people, organizations, courts, case numbers, and addresses.
Citation Analysis
Extract and validate legal citations, find cited authorities.
Privilege Screening
Flag potentially privileged communications before production.
Tone Analysis
Assess the tone and persuasiveness of briefs and filings.
Obligation Extraction
Surface deadlines, duties, and commitments from contracts and orders.
Claim Evolution
Track how claims and arguments change across filings over time.
Argument Structure
Map the logical structure of legal arguments and identify weaknesses.
Exhibit Analysis
Search and analyze exhibit images by visual description.
Pattern Scanning
Regex-based pattern matching across your entire document corpus.
Semantic Search
Natural language search powered by vector embeddings.
Workflow Automation
Chain multiple tools together for complex analysis tasks.
Knowledge Query
Ask natural language questions and get sourced answers.
A Legal Editor Built for the Way You Work
Editing
- Rich text editing with TipTap
- Word (.docx) import and export
- PDF export with headers/footers
- Format Painter for quick styling
- Font size in points (legal standard)
- Heading navigation panel
- Document minimap
Collaboration & AI
- Track changes with accept/reject
- Version history with compare
- AI chat panel with case context
- Auto-suggest from case documents
- Search speed presets (Simple/Medium/Ultra)
Find Anything in Seconds
Sound Suite indexes every document using vector embeddings, enabling semantic search that understands meaning — not just keywords.
Semantic Search
Vector-based search that understands context and meaning across all your documents.
Pattern Matching
Regex search across all documents simultaneously for precise pattern detection.
Exhibit Search
Find exhibit images by describing their contents in natural language.
Multiple Providers
Choose from local Transformers, OpenAI, Anthropic, or Ollama for embeddings.
Reranking
Results reranked for relevance using cross-encoder models for precision.
Instant Results
LanceDB vector store delivers millisecond search across your entire case library.
Set It and Forget It
Point Sound Suite at your case folders. It handles the rest.
File Watcher
Monitors directories for new PDFs. SHA-256 hash deduplication prevents reprocessing.
Text Extraction + OCR
pdfjs-dist extracts text, density analysis detects scans, tesseract.js handles OCR pages.
Exhibit Extraction
Identifies and extracts embedded images, stores in organized exhibit directories.
Text Chunking
HuggingFace tokenizer-based splitting with configurable chunk size and overlap.
Embedding Generation
Multiple providers: local Transformers, OpenAI, Anthropic, or Ollama.
Vector Indexing
LanceDB stores embeddings for fast similarity search. SQLite tracks metadata and status.
Model Context Protocol
Sound Suite exposes all 14 analysis tools via MCP — an open standard for AI tool integration. Connect any MCP-compatible AI assistant to your case intelligence.
// MCP server endpoint
http://localhost:3001/mcp
// Available tools
query_case_knowledge
scan_for_pattern
retrieve_exhibit
// + 11 more analysis tools
See Sound Suite in Action
Download the app and have your first case indexed in minutes.