FuturixAI-and-Quantum-Works/Prism-Legal-OS

Open-source legal AI for contract analysis, legal research, and litigation

What it solves

Legal work is often fragmented across multiple tools for drafting, reviewing, and managing contracts. Prism provides a unified, open-source "Legal OS" that centralizes these workflows into a single connected workspace for legal teams.

How it works

Prism combines a document management system with an integrated AI assistant named Luna. It uses a React frontend and an Express API backed by PostgreSQL for data and Qdrant for vector search and document indexing. Users can upload documents, use AI to extract key terms, compare contracts side-by-side in tabular views, and manage the entire lifecycle from drafting to final approval.

Who it’s for

Legal professionals, law firms, and corporate legal teams who need a collaborative environment to draft, review, and manage contracts with AI assistance.

Highlights

  • AI-Powered Drafting: Research and refine legal documents using the Luna AI assistant with supporting source references.
  • Contract Intelligence: Extract and compare key terms across multiple documents in configurable review tables.
  • AI-Assisted Review: Identify legal risks, missing clauses, and compliance gaps based on custom review playbooks.
  • Template Library: Includes 61 bundled contract and legal-document templates.
  • Collaborative Workflow: Integrated tools for clause editing, version history, and team approvals.
  • Self-Hostable: Can be deployed on private infrastructure using Docker Compose for full control over legal data.

Related

  • Project
    jrz97619761/test-model-thing

    Test‑Model‑Thing (TMT) is a small, MLX‑based language model (≈4.5 M parameters) that experiments with latent‑space prediction, recurrent trace units, byte‑level I/O, continuous streaming, and test‑time training. It runs on Apple Silicon (and Linux) and can be trained or used for chat via a simple CLI. The repo provides training, inference, and CoLA benchmarking scripts, but no pre‑trained weights; users must train their own model.

  • Project
    bibinprathap/VeritasGraph

    VeritasGraph is an open‑source Python framework that replaces vector‑based RAG with a graph‑plus‑tree retrieval engine. It builds a provenance‑rich knowledge graph from documents, lets LLM‑driven agents reason over that graph, and provides a Studio UI for governed agent orchestration (guard‑rails, memory, tool catalog). It runs locally via Ollama or in production with Docker + Neo4j, and offers an MCP server for zero‑trust IDE integration. Key benefits are deterministic multi‑hop retrieval, verifiable citations, explicit abstention, and full on‑prem privacy.

  • Project
    KRR-Oxford/DeepOnto

    A Python package for ontology engineering that integrates deep learning models with formal knowledge representation to support tasks like ontology matching and subsumption prediction.

  • Project
    aws-samples/amazon-bedrock-serverless-prompt-chaining

    A collection of examples and techniques for building scalable, serverless generative AI applications using prompt chaining with AWS Step Functions and Amazon Bedrock Flows.