Legal Case Semantic Search App

---

Project Overview

This project demonstrates how to build a powerful legal case semantic search application using:

• Next.js for full-stack React development
• Pinecone as a vector database for semantic search
• LangChain to parse, chunk, and process PDF legal documents
• Voyage Embeddings to create domain-specific embeddings for legal texts
• TailwindCSS for responsive UI design
• Node.js 20+ for server-side processing

Users can upload and search through legal cases stored as PDFs with natural language queries, enabling efficient retrieval of relevant legal information from large document sets.

---

Key Features

• Programmatic bootstrapping of knowledge bases from PDF documents
• Semantic search optimized for legal text retrieval
• Real-time dynamic suggested search interface
• Automated index creation, embedding, and management for easy onboarding
• Responsive UI using TailwindCSS

---

Tech Stack

Technology	Website
Next.js	https://nextjs.org/
Pinecone	https://app.pinecone.io/
LangChain	https://www.langchain.com/
Voyage Embeddings	https://dash.voyageai.com/
TailwindCSS	https://tailwindcss.com/
Node.js (20+)	https://nodejs.org/en/

---

Getting Started

Prerequisites

• Node.js 20 or higher
• Pinecone API key
• Voyage AI API key
• Environment configured for Next.js development

Installation

git clone https://github.com/BRIKIAchraf/Recherche-s-mantique-juridique-par-l-IA.git
cd legal-case-semantic-search
npm install

Environment Variables

Create a .env file with the following variables:

NEXT_PUBLIC_PINECONE_API_KEY=your_pinecone_api_key
NEXT_PUBLIC_VOYAGE_API_KEY=your_voyage_api_key
NEXT_PUBLIC_PINECONE_ENVIRONMENT=your_pinecone_environment
DATABASE_URL=your_database_connection_string

Running Locally

npm run dev

Open http://localhost:3000 to see the app.

---

Project Structure

• /app - Next.js pages, API routes, and components
• /lib - Helpers for LangChain, Pinecone, and embeddings integration
• /components - UI components styled with TailwindCSS
• /prisma - Database schema and migrations (if applicable)

---

Features Breakdown & Video Timestamps

Timestamp	Feature Description
00:00:00	Introduction
00:00:49	Pinecone Overview
00:05:43	Starter File Setup
00:06:54	Application Overview
00:09:39	Bootstrapping Knowledge Base
00:22:05	Pinecone API Key Setup
00:23:26	Bootstrap + Ingest + Embedding + Search API Implementation
00:50:11	Home Page Development
01:02:50	Testing & Conclusion

---

References and Resources

• Pinecone Docs: https://docs.pinecone.io/
• Voyage AI Dashboard: https://dash.voyageai.com/
• LangChain Docs: https://www.langchain.com/
• Next.js Docs: https://nextjs.org/
• TailwindCSS Docs: https://tailwindcss.com/

---

Contact & Social

For business inquiries: [email protected]

---

License

MIT © 2025 Achraf Briki

---

Subscribe or I turn your VS Code into light mode 😜