atticus

# Atticus – AI-Powered Legal Document Analysis Platform

Atticus is a full-stack AI platform designed for legal teams to manage case documents, search across large collections of files, and receive AI-generated answers with source citations.

The system supports secure case-based access control, asynchronous document ingestion, hybrid search, and conversational AI. Users can upload legal documents in multiple formats (PDF, DOCX, TXT, EML), which are automatically processed, indexed, and made searchable through a Retrieval-Augmented Generation (RAG) pipeline.

Key Features:

* Role-based authentication and authorization (Admin / Lawyer)

* Case-scoped document management

* Asynchronous document processing pipeline

* Hybrid retrieval using vector search and keyword search

* AI-powered chat with source citations

* Real-time streaming responses

* Secure document access controls

* Full Dockerized deployment environment

Technologies:

* Frontend: Next.js, React, TypeScript

* Backend: FastAPI, Python

* Database: PostgreSQL

* Vector Database: Qdrant

* Search Engine: Elasticsearch

* Queue & Cache: Redis, Celery

* AI Integration: LLM-powered Retrieval-Augmented Generation (RAG)

* Deployment: Docker Compose

This project demonstrates full-stack development, backend architecture, AI integration, search systems, asynchronous processing, and production-oriented software engineering practices.