AI Portfolio

Dr. Youssef Mohamed

Social AI Researcher & Entrepreneur

PhD in Social AI from KTH Royal Institute of Technology. Specializing in Human-Robot Interaction, Affective Computing, Computer Vision, and Applied Machine Learning. Building adaptive AI systems for real-time human state detection and multi-modal analysis.

youssefezzat12@gmail.com
+46 790489221
LinkedIn Profile

AI Solutions Portfolio

Cutting-edge AI projects demonstrating expertise in computer vision, natural language processing, and cloud-native architectures.

Computer Vision • GCP

Vizioneer

AI-Powered Camera Analytics for Dark Stores & Cloud Kitchens

Visit Site

Vizioneer transforms existing camera infrastructure into a real-time operational command center for rapid commerce. The system uses state-of-the-art Vision Language Models (VLMs) and machine vision algorithms to track orders, optimize picking routes, and detect bottlenecks in dark store and cloud kitchen environments—all from existing cameras with zero additional hardware.

Technology Stack

Vision Language Models (VLMs) Multiple Vision Models Machine Vision RTSP Streaming WebSockets Real-time Processing Google Cloud Platform Object Detection Tracking Algorithms

Key Features

Smart Picking Routes
Real-Time Inventory
Order Verification
Bottleneck Detection
40%
Faster Picks
99.5%
Accuracy
2min
Setup Time
Vizioneer Dashboard - 3D Visualization
AI-Powered Analytics Dashboard
Vizioneer Dark Store Solutions
Dark Store Optimization
Vizioneer Platform Features
Platform Architecture
LLMs • Behavioral AI • GCP

Pokamind

AI-Native Adaptive Learning & Leadership Development Platform

Visit Site

Pokamind is an AI-powered skill-training platform that delivers personalized leadership development through realistic role-play scenarios and multi-modal behavioral analysis. The system analyzes body language, voice patterns, and communication keywords in real-time to provide objective insights and accelerate skill development. Built on enterprise-grade GCP architecture with Kubernetes orchestration, achieving 99.9% uptime.

Technology Stack

Large Language Models Explainable AI (SHAP/LIME) RAG Architecture LoRA Fine-tuning Communication Analysis Body Language Detection Voice Analysis Vision Models Google Cloud Platform Kubernetes OAuth 2.0

AI Components

Conversational AI Engine
Explainable AI Feedback
Adaptive Personalization
Communication Pattern Analysis
4x
ROI
68%
Faster Development
92%
Satisfaction
99.9%
Uptime
Pokamind Platform Overview
AI-Native Learning Platform
Pokamind Enterprise Features
Enterprise Security & Scalability
Computer Vision • Tracking AI • GCP

HighlightsHub

Automatic Goal Detection & Sports Analytics Platform

Visit Site

HighlightsHub leverages state-of-the-art computer vision and object tracking AI to automatically detect goals in casual football games and generate professional highlight reels. The system uses fine-tuned vision models for precise event detection, player identification, and action tracking. Built on a full Google Cloud Platform architecture with Firebase integration for real-time processing and seamless user experience.

Technology Stack

Fine-tuned Vision Models Object Detection (YOLO/Detectron2) Object Tracking AI Action Recognition Player Identification Event Detection Google Cloud Platform Firebase Real-time Processing Video Analytics

Core Features

Automatic Goal Detection
Player Identification
Highlight Generation
Match Statistics
AI
Auto Detection
Real-time
Processing
Cloud
Architecture
HighlightsHub Platform Interface
Platform Features
NLP • Social Robotics • HRI

Furhat Robot - Arabic NLP System

Social Robot for Cultural Engagement in NEOM

Visit Furhat Robotics

Developed an advanced Natural Language Processing system for Furhat Robotics' social humanoid robot, specifically adapted for Arabic language and Middle Eastern cultural contexts. The system was deployed at the NEOM Exhibit Center in Saudi Arabia, where the robot conducted interactive surveys and engaged with Saudi visitors in natural, culturally appropriate Arabic conversations. This project represents a significant advancement in cross-cultural human-robot interaction, enabling seamless bilingual communication and culturally-aware social engagement.

Technology Stack

Kotlin Java Natural Language Processing Arabic NLP Dialectical Arabic Speech Recognition Speech Synthesis ROS (Robot Operating System) Cultural Adaptation AI Conversational AI Social Robotics

Key Achievements

Arabic Language Processing
Cultural Context Adaptation
Natural Conversation Flow
NEOM Deployment
NEOM
Exhibit Center
Arabic
NLP Engine
HRI
Social Robotics

About NEOM & Furhat Robotics

NEOM is Saudi Arabia's $500 billion futuristic smart city project featuring AI-driven infrastructure, robotics integration, and autonomous systems as part of Saudi Vision 2030. Furhat Robotics is a Swedish startup creating social robots with human-like facial expressions and natural conversation capabilities, used in research institutions and commercial deployments worldwide for customer service, healthcare, education, and recruitment.

Furhat Social Robot
Furhat Social Robot with Human-like Expressions

Research & Publications

Explore my academic contributions in Human-Robot Interaction, Affective Computing, and Social AI through published research and open-source frameworks.

View Publications About Me