02 / AI / LegalTechIndependent project / Tharun

Juris

Legal answers, grounded in the document.

A document-intelligence workspace that brings contracts, structured analysis and source-grounded questions into one interface.

Visit Live Website

Product in focus / JurisDesign → Engineering → Delivery

Juris introduction with monumental legal intelligence typography and a document archive graphic. — 02 / From documents to intelligence

01 / My role: Full-Stack Developer · Product Engineering · AI Integration
02 / Core technologies: React / FastAPI / OCR / OpenAI
03 / Engineering focus: Document intelligence / Source-grounded answers

01 / Overview

Keep the answer close to its source.

An AI-powered legal intelligence platform that transforms contracts into structured, searchable and source-grounded insights.

The problem

Contracts spread important information across clauses, obligations, dates and payments. Juris brings those details into a searchable workspace, with structured analysis and document-grounded questions.

What I built

I built the full-stack product and AI integration, connecting a React interface with a Python/FastAPI backend and document-intelligence capabilities.

02 / Product & experience

Two contexts. Clear boundaries.

Document chat works with the uploaded agreement. Juris Compass provides a separate place to explore laws, concepts and procedures in a selected jurisdiction. The interface distinguishes questions about a document from broader legal research.

Source → UnderstandingProduct logic / A visual study

01
OCR & structure
02
Clauses & summaries
03
Answers with sources

Begin with the uploaded document. OCR and structure make its content available to the analysis.

Document-grounded answers stay connected to the source. Juris Compass is a separate research context.

Document intelligence

Explore structured summaries, clauses and risk analysis from the document workspace.

Source-grounded questions

Ask about obligations, dates or terms, with sources included in the response.

Juris Compass

Research legal concepts and procedures beyond an individual document.

Compass / Research beyond the document.

03 / System & engineering

From document handling to grounded analysis.

Juris uses Tesseract OCR, embeddings and OpenAI GPT models alongside a FastAPI backend. Supabase, PostgreSQL and Cloudflare R2 form the data and storage stack.

01Interface: React / Vite / JavaScript
02Analysis: Python / FastAPI / Tesseract OCR / embeddings / OpenAI GPT models
03Data & storage: Supabase / PostgreSQL / Cloudflare R2
04Delivery: Vercel / Render

04 / Selected screen

The document stays in view.

The document workspace keeps the file, analysis areas and suggested questions together. A dedicated question field makes the scope explicit: ask about this agreement, with sources included.

Juris document chat workspace with a processed rental agreement, source-grounded questions and analysis navigation. — 01 / Document workspace

05 / Delivered

A document workspace connecting OCR, structured analysis and questions grounded in the source.

I built the full-stack product and AI integration, connecting a React interface with a Python/FastAPI backend and document-intelligence capabilities.

Explore the live project

Juris / React / FastAPI / OCR / OpenAI