Skip to main content

Artificial intelligence

AI Intelligence Project Management System

AI-powered enterprise project and task management platform with intelligent task generation and smart assignment.

Evidence
Verified work
Role
Project manager&Ai features
Timeline
Jan 2026
Category
Artificial intelligence

Overview

An advanced AI-powered enterprise project and task management platform that combines intelligent task generation, smart task assignment algorithms, and comprehensive project management capabilities. The system features multi-company support, role-based access control (Super Admin, Admin, Project Manager, Developer, Client), real-time collaboration, and built-in delay tracking with AI-driven insights.

Problem

Project work across several companies and five roles (Super Admin, Admin, Project Manager, Developer, Client) turns on two recurring decisions: how a goal becomes concrete tasks, and who should do each task without creating time conflicts. The platform brings both decisions — together with delay tracking — into one system.

Constraints

Multi-company tenancy with role-based access control; real-time collaboration; delivery across a Next.js web client, a Flutter mobile client and a Tauri desktop build; and an assignment step that must respect existing schedules.

Approach

We designed and developed an AI-powered enterprise project management platform that combines intelligent task generation, automated task assignment, project management, and real-time collaboration in a unified system. The platform uses AI-driven logic to help generate and organize tasks, intelligently assign work based on project and team requirements, and provide insights into delays and project progress. A multi-company architecture with role-based access control enables secure management across Super Admin, Admin, Project Manager, Developer, and Client roles. The solution also provides centralized project and task management, real-time collaboration, delay tracking, and AI-driven insights, creating a scalable foundation for organizations seeking to improve project efficiency, task automation, team coordination, and enterprise workflow management.

Architecture

Built with a full-stack architecture using Node.js/Express backend with MongoDB, Next.js frontend, and Flutter mobile application. The AI components include task generation using multiple LLMs (Groq API, Mistral 7B, Qwen 2.5), a custom neural network with FAISS retrieval system, a greedy algorithm for optimal task assignment with time conflict detection, and an expert system for task classification.

Key technologies: Node.js, Express.js, MongoDB, Next.js, React, Flutter, Tailwind CSS, FastAPI, JWT Authentication, Argon2 hashing, Helmet, CORS, Rate Limiting, Chart.js, Framer Motion, Swiper, and Tauri for desktop deployment.

Intelligence layer

The intelligence layer is split into four stages. Task generation uses several large language models (Groq API, Mistral 7B, Qwen 2.5) to propose candidate tasks from project context. A custom neural network with a FAISS retrieval system brings related prior work into that context. An expert system classifies tasks. A greedy algorithm then assigns tasks to people, detecting time conflicts before a task is placed.

Engineering decisions

Generation and assignment are deliberately different kinds of computation: open-ended decomposition is delegated to language models, while the step that must be correct — who does what, and when — is an explicit algorithm whose output can be checked. Classification by an expert system keeps the categorisation rules inspectable rather than learned implicitly.

The platform side uses JWT authentication with Argon2 password hashing, Helmet, CORS and rate limiting — security controls that sit in front of every AI-backed endpoint.

Results

The platform delivers a centralized, AI-powered environment for managing projects, tasks, teams, and organizational workflows. Intelligent task generation and assignment reduce manual project coordination, while AI-driven insights provide greater visibility into workload, progress, and potential delays. The multi-company architecture and role-based access control enable organizations to manage projects securely across different teams and user roles. Real-time collaboration improves communication between project stakeholders, while built-in delay tracking helps teams identify issues earlier and respond proactively. The result is a scalable enterprise project management platform that brings together AI-powered task automation, intelligent work allocation, real-time collaboration, project monitoring, and data-driven insights in a single workflow.

Media

AI-powered project management dashboard showing projects, active tasks, team members, project progress, and AI insights
AI project management Kanban board showing product development tasks across workflow stages
AI assistant interface for intelligent task generation, project planning, and task assignment
AI project management analytics dashboard showing project completion, team productivity, task completion, and performance metrics
Team management dashboard showing project roles, active tasks, workload distribution, and team member status

Software engineering

Dr. Elias Dahdal Clinic — Clinic Management System

A multi-department clinic management system in production: reception, laser, dermatology and dental workflows, billing and accounting in two currencies, payroll, inventory, audit and a patient portal.

Stack

  • React
  • TypeScript
  • Three.js / WebGL
  • Node.js
  • Express.js
  • MongoDB
  • REST API design
  • Authentication & session security
  • Authorization (RBAC)
  • Internationalization & RTL
  • CI/CD
  • Documented: Problem, Approach, Architecture, Results
Dr. Elias Dahdal clinic system — the public sign-in page for staff and patients (Arabic, right-to-left)

Artificial intelligence

Concept system

Enterprise RAG Platform

Concept system — a retrieval-augmented generation platform: document ingestion, chunking, embeddings, hybrid retrieval, re-ranking and grounded answers with citations and evaluation.

Stack

  • Retrieval-augmented generation
  • Large language models
  • FAISS vector retrieval
  • Asynchronous processing & queues
  • PostgreSQL
  • Documented: Problem, Architecture, Intelligence layer, Design objective

Software engineering

Tavola — Multi-Tenant Restaurant Reservation Platform

A multi-tenant SaaS for restaurant reservations, tables, branches and staff: NestJS backend with real-time updates, a Flutter customer app, a React restaurant dashboard and a platform-owner console.

Stack

  • NestJS
  • TypeScript
  • PostgreSQL
  • Prisma
  • Redis
  • BullMQ
  • Real-time systems (WebSockets)
  • Multi-tenant SaaS architecture
  • Authentication & session security
  • Authorization (RBAC)
  • Object storage (S3 / MinIO)
  • Docker
  • Nginx
  • Flutter
  • Dart
  • React
  • Tailwind CSS
  • Internationalization & RTL
  • System design & ADRs
  • Automated testing
  • Source
  • Documented: Problem, Approach, Architecture, Results