LexQuery — Indian Legal Intelligence System

Ask questions across IPC sections, Supreme Court judgments, and contracts — powered by RAG, LLaMA3, and ChromaDB.

---

Problem

Legal research in India is slow, expensive, and inaccessible. Finding the right IPC section, understanding a court judgment, or spotting risky contract clauses requires hours of manual reading. LexQuery makes it conversational.

---

What It Does

• IPC Q&A — "What is the punishment for culpable homicide?" → cited answer from IPC_1860
• Judgment summarization — "Summarize the Maneka Gandhi judgment" → plain-English answer with source
• Contract analysis — "What does the indemnity clause say?" → answer grounded in the uploaded contract
• Filtered queries — search only IPC, only judgments, or only contracts

---

Tech Stack

Component	Tool
PDF extraction	pdfplumber
Chunking	LangChain RecursiveCharacterTextSplitter
Embeddings	HuggingFace all-MiniLM-L6-v2
Vector store	ChromaDB
LLM	Groq LLaMA3-8b
Orchestration	LangChain

---

Project Structure

lexquery/
├── data/
│   ├── ipc/             # IPC_1860.pdf
│   ├── judgments/       # landmark SC judgment PDFs
│   └── contracts/       # sample contract PDFs
├── src/
│   ├── ingest.py        # load → clean → chunk → embed → store
│   ├── retriever.py     # query ChromaDB with optional filters
│   └── chain.py         # Groq LLaMA3 QA chain
├── main.py              # CLI entry point
├── config.yaml          # all settings
├── .env.example         # API key template
└── requirements.txt

---

Setup

# 1. Clone
git clone https://github.com/yourusername/lexquery.git
cd lexquery

# 2. Install dependencies
pip install -r requirements.txt

# 3. Set up API keys
cp .env.example .env
# Add your GROQ_API_KEY and HF_TOKEN to .env

# 4. Add PDFs to data/ folders (see Data Sources below)

# 5. Ingest
python main.py --ingest

---

Usage

# Query across all documents
python main.py --query "What is the punishment for murder under IPC?"

# Filter by document type
python main.py --query "What does Section 420 say?" --doc_type ipc
python main.py --query "Summarize the Maneka Gandhi judgment" --doc_type judgment
python main.py --query "What are the indemnity terms?" --doc_type contract

---

Data Sources

All data is publicly available and free to download:

Type	Source	URL
IPC 1860	India Code (Govt. of India)	https://indiacode.nic.in
SC Judgments	Indian Kanoon	https://indiankanoon.org
SC Judgments	Supreme Court of India	https://sci.gov.in
Sample Contracts	TLDR Legal	https://tldrlegal.com

---

Configuration

All settings are in config.yaml — no hardcoded values in code:

embedding:
  model: "sentence-transformers/all-MiniLM-L6-v2"
llm:
  model: "llama3-8b-8192"
chunking:
  chunk_size: 800
  chunk_overlap: 100
retrieval:
  top_k: 5

---

Author

Swastik — CS Engineering Student, building toward AI Engineering.