CASE STUDYLegal application + AI

Justicelib

AI-assisted legal platform

5
Technologies

Laravel · React · OpenAI API

4
Key features

Advanced AI integration

4
Documented screens

Comment ça marche

Live
Status

Legal application + AI

CASE STUDY

The context

Justicelib is a French platform that makes common legal procedures (formal demand letters for unpaid rent, consumer disputes, administrative requests) accessible to individuals and small businesses. The target audience: anyone paying too much for a lawyer to draft acts that the law lets you handle yourself, provided you use the right template, the right code articles, and the right tone. The product bet was to use an LLM to generate these custom documents from a 5-minute form, and deliver them ready to sign.

The challenge

Three constraints in tension. First, legal reliability: an AI-generated draft citing the wrong Consumer Code article is unusable, so we had to combine LLM output with locked templates. Second, traceability: every generated document must be auditable (which template, which fields, which version of the legal base). Third, UX: users are non-lawyers stressed by a conflict. The form had to be short, reassuring, and never ask for information the client wouldn't know where to find.

Our approach

Laravel backend with a business layer orchestrating the OpenAI call: the LLM doesn't generate the full document, it fills legal templates pre-validated by a partner lawyer. Each template has a version (Law 2024-x), an identifier, a required-fields checklist. React + Tailwind frontend for the form funnel, mobile-first by default because most users arrive from a smartphone after searching "mise en demeure loyer" on Google. MySQL stores submissions and generated document versions, with secure authentication so users can retrieve their history.

The outcome

The platform covers around ten legal procedures today (unpaid rent, refund refusal, payment-deferral request, etc.). Generated documents come as print-ready PDFs with recommended attachments. The generation funnel runs under 6 minutes median including the user's proofreading step. On the ranking side, the site captures "mise en demeure [topic]" queries historically held by generalist law-firm portals.

KEY PAGES

The product in detail

A glimpse of the screens that make the product valuable: features, user journey, content.

justicelib.com/comment-ca-marche
Justicelib · Comment ça marche
Comment ça marcheView
justicelib.com/mise-en-demeure-loyer-impaye
Justicelib · Mise en demeure loyer
Mise en demeure loyerView
justicelib.com/mise-en-demeure-consommation
Justicelib · Mise en demeure conso
Mise en demeure consoView
justicelib.com/faq
Justicelib · FAQ
FAQView

Key features

Advanced AI integration
Intuitive user interface
Secure authentication system
Admin dashboard

Technologies used

LaravelReactOpenAI APIMySQLTailwind CSS

Have a similar project?

We can build a custom solution for your business. Let's talk.

Issam Rougui, fondateur de ROUGUITECH

MEET

Issam
Rougui.

Founder & Lead Developer · ROUGUITECH

Full Stack developer with 5+ years of experience, specialized in Laravel, React, WordPress, and SEO. I founded ROUGUITECH to help companies in Morocco and internationally turn their digital presence into a growth engine.

Find me on
WhatsApp