|
|
| <div align="center"> |
| <img src="https://cdn-avatars.huggingface.co/v1/production/uploads/69e9f4d92a09c772420b42d0/ExIuqQi9htcT0oFbZAgUu.png" width="140" alt="JudicialMind logo" /> |
|
|
| <h1>JudicialMind</h1> |
|
|
| <p><strong>The AI-Native Legal Intelligence System</strong><br/> |
| Powering the world's most ambitious law firms and enterprise legal departments.</p> |
|
|
| <p> |
| <a href="https://judicialmind.ai"><img alt="Website" src="https://img.shields.io/badge/website-judicialmind.ai-0ea5e9"></a> |
| <a href="https://huggingface.co/datasets/judicialmind/legal-training-dataset"><img alt="Flagship dataset" src="https://img.shields.io/badge/dataset-legal--training-ffd21e"></a> |
| <img alt="License" src="https://img.shields.io/badge/research-open--access-success"> |
| </p> |
| </div> |
| |
| --- |
|
|
| ## 🧭 Who we are |
|
|
| JudicialMind builds an orchestrated system of specialized AI agents that work like a legal team — every agent with a domain expertise, composed into directed workflows that run end-to-end legal operations: intake, research, drafting, negotiation, dispute resolution. |
|
|
| We serve: |
| - **AmLaw firms** — matter-level intelligence and scalable research capacity |
| - **In-house legal teams** — contract, risk, and compliance workspaces |
| - **Solo & boutique practices** — affordable, agent-led practice leverage |
| - **Courts & ADR bodies** — dispute triage and caseflow intelligence |
|
|
| --- |
|
|
| ## 🧠 The Platform |
|
|
| | Product | What it does | |
| | --- | --- | |
| | **Intelligence** | Legal knowledge graphs, case-law reasoning, jurisdiction-aware retrieval | |
| | **Workspace** | Agent-native drafting, review, and matter orchestration | |
| | **Disputes** | ODR / ADR workflow automation and dispute resolution agents | |
|
|
| --- |
|
|
| ## 📦 Open artifacts on Hugging Face |
|
|
| We publish the datasets and components that power our research openly wherever we can. |
|
|
| ### Datasets |
| - [**`judicialmind/legal-training-dataset`**](https://huggingface.co/datasets/judicialmind/legal-training-dataset) — 3.69M annotated query-passage pairs across **35 languages** for training legal IR, QA, and semantic-search models. Rich metadata per row: `query_type`, `legal_domain`, `difficulty`, `jurisdiction`. ~2.6 GB, 264 parquet shards, A/B/C file-level split for clean train/val/test partitioning. |
|
|
| ### Models |
| _Coming soon — we are releasing family of fine-tuned retrieval, reranking, and legal-reasoning models trained on our corpus._ |
|
|
| ### Spaces |
| _Coming soon — interactive demos for legal retrieval, agent workflows, and dispute-resolution tooling._ |
|
|
| --- |
|
|
| ## 🔬 Research focus |
|
|
| - Multilingual legal retrieval at scale (35+ jurisdictions) |
| - Reasoning over structured legal knowledge graphs |
| - Agent composition for multi-step legal workflows |
| - Evaluation frameworks for factuality, citation integrity, and jurisdictional grounding |
|
|
| --- |
|
|
| ## 🤝 Connect |
|
|
| - 🌐 Website — https://judicialmind.ai |
| - 📧 Contact — [surya@judicialmind.ai](mailto:surya@judicialmind.ai) |
| - 🧑⚖️ Early access — [judicialmind.ai/waitlist](https://judicialmind.ai/waitlist) |
|
|
| <sub>© JudicialMind. Built with care for the practice of law.</sub> |
|
|
