Case file 01 — Co-founder & Lead Engineer

FASL

Multi-agent AI that drafts Moroccan legal appeals end-to-end.

58.4KSearch impressions in 28 days · Google Search Console

Article 01 — The problem

Drafting an appeal memo in Morocco means hours of case analysis, digging through legal codes, and careful Arabic legal writing — with no tooling built for how Moroccan lawyers actually work.

Article 02 — The procedure

  1. §1Built a multi-agent pipeline that carries a case from analysis to a polished legal memo: case analysis → document retrieval → draft generation → final review.
  2. §2Engineered a RAG pipeline over Moroccan legal codes so every citation is grounded in real law, minimizing hallucination in a domain where mistakes are expensive.
  3. §3Kept lawyers in control with a human-in-the-loop rich-text editor for reviewing and refining every draft before it leaves the building.
  4. §4Designed the product Arabic-first with full RTL support, because the legal system runs in Arabic — not in translation.

Due process — how a case moves through the agents

01

Case analysis

An agent reads the case file and extracts the legal issues at stake.

02

Retrieval

RAG over Moroccan legal codes surfaces the articles and precedents that apply.

03

Drafting

A drafting agent composes the appeal memo, grounded in the retrieved law.

04

Human review

The lawyer refines the draft in a purpose-built Arabic legal editor.

Exhibits — entered into evidence

Exhibit A — the platform

FASL landing page — Arabic legal AI platform for drafting appeal memos

Exhibit B — in hand

FASL on mobile

The verdict

  • Live in production at fasl.ma with lawyers reaching out inbound.
  • 58.4K organic search impressions and 1.45K clicks in the last 28 days — 95.3K impressions over 3 months at average position 6.2 (Google Search Console), zero ad spend.
  • Next.js
  • TypeScript
  • Claude & Gemini
  • RAG
  • PostgreSQL
  • Prisma
  • n8n
  • TipTap