Skip to content

Repository files navigation

LexCore AI

Open-source legal case management with AI-assisted PDF analysis.

Source code · Issues · Contributing · MIT license

LexCore AI brings clients, legal cases, documents, and reports into one application. Built with Rust, Tuono, React, and PostgreSQL, it connects to Ollama to answer questions about uploaded PDFs.

The project is a development prototype. Its interface and domain identifiers use Portuguese; this documentation uses English.

About

The canonical repository is felpzw/LexCore-AI. The project was previously called Jurídico IA, with the repository name Projeto-Banco-de-Dados. Development continues here as LexCore AI.

Contributions branch from and target develop; master is reserved for releases. See repository identity and GitHub About for canonical links, attribution, and repository description and topic values.

Capabilities

  • Client management: create, view, update, and delete individual and corporate client records.
  • Case management: associate cases with clients, lawyers, statuses, courts, and categories.
  • Document management: upload, view, download, update, and delete documents linked to cases.
  • AI-assisted PDF analysis: select an available Ollama model and ask questions using text extracted from a stored PDF.
  • Reports: visualize documents by client and case, cases by lawyer and status, and hearings by client and lawyer.
  • Development utilities: initialize the database schema, load sample records, and reset a disposable database.

Technology

Layer Implementation
Application framework Tuono 0.19.7
Backend Rust, Tokio, tokio-postgres, Reqwest
Frontend React 19, TypeScript, Recharts
Database PostgreSQL 16
AI inference Ollama
Local services Docker Compose

Get started

Clone the canonical repository and switch to the development branch:

git clone https://github.com/felpzw/LexCore-AI.git
cd LexCore-AI
git switch develop

Follow the setup and environment guide for prerequisites, configuration, model installation, and database initialization.

After completing the prerequisites and configuring var.env, run from the repository root:

npm ci
docker compose up -d
tuono dev

Open http://localhost:3000. On a fresh database, visit /configuracoes to initialize the schema and load sample records before using the management pages.

The database reset action drops every table in the connected database's public schema. Use it only with a disposable development database.

Documentation

Guide Contents
Documentation index Reading order and documentation maintenance
Repository identity and About Canonical links, project history, GitHub metadata, and attribution
Project overview Architecture, modules, data flow, and current limitations
Setup and environment Local installation, configuration, and troubleshooting
Git workflow Branching, pull requests, releases, and hotfixes
Commit standards Commit format, types, scopes, and examples

Repository layout

docs/                  Project and contributor documentation
src/app.rs             Application startup and environment loading
src/lib.rs             Shared database and query helpers
src/routes/            React pages and Rust route handlers
src/routes/api/        Backend API endpoints
src/components/        Reusable React components
src/styles/            Application styles
docker-compose.yml     PostgreSQL and Ollama services
var.env                Local runtime configuration (currently tracked)

Contributing

Read the contributing guide, Git workflow, and commit standards. Open regular pull requests against develop. Keep changes focused, update the relevant documentation, and describe the checks actually performed in each pull request.

License and acknowledgments

LexCore AI is licensed under the MIT License.

The application builds on open-source technologies including Tuono, React, PostgreSQL, Ollama, and Recharts. Dependencies and downloaded models retain their own licenses; see the attribution notes.

About

No description, website, or topics provided.

Resources

Contributing

Stars

0 stars

Watchers

0 watching

Forks

Releases

Packages

Contributors

Languages