Skip to content
Tasklet

Al-Shalawi Law · Saudi law firm · bilingual legal operations

One system for cases, courts, contracts and counsel

A bilingual Arabic/English practice platform that replaces spreadsheets and WhatsApp groups with cases, clients, court sessions, AI drafting, WhatsApp automation and multi-channel alerts.

Full-stack delivery
Al-Shalawi Law
Al-Shalawi dashboard with open cases, sessions, tasks and upcoming court dates

What we did, plainly

We designed the architecture and built the backend, frontend, CI/CD, AI legal chat and WhatsApp integration. The platform runs on Google Cloud Run with staging and production environments. Screens below recreate the production layout with invented seed data; sensitive fields are blurred.

18
feature modules in the backend
35
database migrations
35
application pages
1,000+
translation keys per language

Counted from the codebase in November 2026. Screens recreate the production UI layout with invented data; staging was unreachable at capture time.

The problem

A law firm’s day-to-day work was split across spreadsheets, WhatsApp groups and disconnected reminders. Court dates, contracts, POAs and deadlines lived in different places, and bilingual Arabic/English work made the patchwork harder.

Staff needed one system that understood Saudi calendars (Hijri and Gregorian), attorney roles, and the messaging channels the firm already uses — without turning legal judgment into a black-box chatbot.

Why off-the-shelf was not enough

Generic practice tools do not model Hijri court sessions, Najiz-style WhatsApp notifications, bilingual RTL operations and firm-specific drafting references in one place. The firm needed software shaped around how it already works.

What the platform does

Al-Shalawi is a role-aware legal operations platform: clients, cases, contracts, court sessions, POAs, tasks, attendance, files and notifications in one bilingual application.

On top of that core, legal chat drafts documents from firm references with structured outputs and export, a WhatsApp bot handles signed court notifications with human confirmation, and a generic alerting engine reminds people over email, WhatsApp, SMS or webhook.

How it works

How the platform works, step by step

  1. 01 / Dashboard

    Start from what needs attention

    Open cases, sessions this week, tasks and contracts approaching expiry sit on one screen, scoped to the attorney or firm-wide for admins.

    LiveOur work
    Al-Shalawi Law
    Dashboard KPIs and upcoming court sessions with Hijri dates
  2. 02 / Cases

    Cases with stages, types and owners

    Cases carry type, status, lifecycle stage and lead attorney, with multilingual names and search so the file is findable in either language.

    LiveOur work
    Al-Shalawi Law
    Cases table with type, stage, attorney and status
  3. 03 / Court sessions

    Hijri and Gregorian side by side

    Court sessions store both calendars, Saudi timezone, attorney filters and status so nothing depends on a personal reminder.

    LiveOur work
    Al-Shalawi Law
    Court sessions list with Gregorian and Hijri columns
  4. 04 / Legal chat

    Drafting grounded on firm references

    Attorneys draft POAs, memos and contracts in Arabic or English with token-budgeted context, structured outputs and DOCX/PDF export. The assistant proposes document type and category before references load.

    LiveOur work
    Al-Shalawi Law
    Legal chat session proposing a power of attorney draft
  5. 05 / Alerts

    Reminders that reach the right channel

    A generic alerting engine materialises schedules from configs, sends over email, WhatsApp, SMS or webhook, and keeps an audit log. Credentials are encrypted at rest.

    LiveOur work
    Al-Shalawi Law
    Alert configs and recent send log across email and WhatsApp

Interface screens recreate the production platform layout with invented seed data. Sensitive case, attorney and recipient fields are blurred. Live marketing site: alshalawilaw.com.

Capabilities

What exists, and what does not.

Every capability carries a status. Nothing is listed as live unless it is running in production.

Live
Running in the client's production environment.
Built
Implemented in the codebase, but switched off for production users or waiting on configuration.
Could build
Not built. Something we could build for you, shown so you can see where the work could go.
  • Clients, cases, contracts and POAs

    Core practice records with multilingual fields and role scoping.

    Live
  • Court sessions with Hijri and Gregorian

    Saudi timezone, attorney filters, dual-calendar display.

    Live
  • Tasks, notes, files and notifications

    Day-to-day work beside the matter file.

    Live
  • Dashboard KPIs

    Case distribution, upcoming sessions, tasks and contract expirations.

    Live
  • Legal chat with structured drafting

    OpenAI models, token budgets, firm references, DOCX/PDF export.

    Live
  • WhatsApp Business bot

    Signed webhooks, Najiz-aware session extraction, human confirm buttons.

    Live
  • Multi-channel alerting engine

    Email, WhatsApp, SMS, webhook; encrypted channel credentials.

    Live
  • GPS attendance and office geofencing

    Check-in against configured office places.

    Live
  • Google Sheets court-session sync

    Dual-tab English/Arabic mirror of sessions.

    Live
  • Bilingual RTL/LTR application

    Locale-prefixed routes, 1,000+ keys per language.

    Live
  • Cloud Run CI/CD with Secret Manager

    Staging and production, self-bootstrapping pipelines.

    Live
  • Vector RAG over the firm corpus

    pgvector is wired; full retrieval-augmented chat is a natural extension.

    Could build

Our work Everything not marked as “Could build” was engineered by us.

Operational value

What it changes day to day.

Described from how the system works. We do not quote savings or percentages we cannot evidence.

One operational backbone
Cases, courts, contracts and people stop living in separate chats and sheets.
Bilingual by design
Arabic RTL and English LTR are first-class at the URL, content and date layer — not a bolted-on translation.
AI that stays inside professional bounds
Drafting is grounded on firm references, structured, exportable and capped — with people confirming sensitive WhatsApp actions.
Reminders that actually send
Configurable multi-channel alerts with an audit trail replace manual chasing before deadlines.

Built with

  • Next.js and React
  • TypeScript
  • Ant Design
  • FastAPI on Python
  • PostgreSQL with pgvector
  • OpenAI models
  • WhatsApp Business Cloud API
  • Google Cloud Run, Cloud Build and Secret Manager
  • Sentry and OpenTelemetry

Next

Page7

AI social media operations platform

Read the Page7 case study

Running a professional practice on chats and sheets?

Tell us how cases, courts, contracts and reminders work today. We will tell you what a custom platform could take on — and what should stay with people.