Dirah’s Intranet Portal
2 MonthsProject duration
10 PagesTotal Pages
TourismIndustry
Next.js
TypeScript
AWS
MongoDB
Overview
Beyond the public-facing Dirah Destination experience, Greenmint Labs built a secure internal operations portal for Al Dirah's staff, a bilingual back-office platform for managing property listings, internal documentation, and day-to-day team communication. The portal is built around a full service management system, so every request an employee raises is created, routed, fulfilled, and closed out inside one connected workflow rather than over email or spreadsheets, with AI layered into the moments that used to take the most manual effort.
Core Capabilities
Service Management System
A Request Centre gives every employee one place to browse and submit service requests for document delivery, stationery, accident notifications, authorisation requests, and more. Employees are able to track them from submission through fulfilment. Each request carries a live status (Pending, Reviewing, Resolved), a full activity timeline, and a request ID, so nothing gets lost in someone's inbox.
Role-Based Access Control (Fulfillers & Approvers)
Access is scoped by role. Requesters see "My Requests"; people in an approval chain get a dedicated "My Approvals" queue for anything awaiting their sign-off; and the teams who actually complete the work operate from a separate Fulfillers view, so a stationery request and a document-delivery request can be routed to two entirely different back-office teams without either seeing the other's queue.
Form Builder & Workflow Builder
Each request type is defined by a configurable form and a matching approval/fulfillment workflow, so Al Dirah's admins can stand up a new request type- new fields, new approval steps, new routing for any department without a development cycle.
AI-Powered Features
AI-Powered Semantic Search & Virtual Assistant
A conversational search bar replaced the old keyword-only search across the portal. Staff can ask plain-language questions ("how do I request annual leave?") and the assistant matches intent rather than literal text, surfacing the right form, policy, or property record directly instead of a page of unrelated results. It runs as a persistent chat widget available from every screen in the portal.
AI-Assisted Multilingual Content
Because Al Dirah publishes every internal announcement, listing update, and policy in both Arabic and English, content teams previously wrote and maintained two versions of everything by hand. A translation layer now drafts the counterpart language automatically the moment new content is published, tuned to Al Dirah's own terminology, and routes it to an editor for a quick review before it goes live, cutting duplicate-authoring work while keeping both language versions consistent.
AI Document Intelligence for the Document Library
The portal's contract templates, listing specifications, and internal guides are now indexed by a retrieval-based assistant that answers direct questions with a citation back to the exact source document and section, instead of requiring staff to open and search entire PDFs by hand.
Outcome
The result is a bilingual internal platform that mirrors the polish of Al Dirah's public sites while solving a very different problem: giving a real estate team one governed system for every internal request, with the right people — requester, approver, fulfiller, seeing exactly what they need to at each step, and AI removing the busywork around search, translation, and documentation along the way.
More projects
Let’s build something that matters
Have an idea, a problem to solve, or a product you want to improve?
