I’m Ammar, an AI Consultant & Engineer, Machine Learning Specialist, and University Lecturer.
I am Ammar Mohanna, an AI Consultant with extensive experience in both academia and industry, currently serving as a Lecturer at the American University of Beirut (AUB) and a Senior AI Consultant at EDT&Partners in Spain. My journey in artificial intelligence began with a Masters in Software Engineering from the School of Engineering of Beirut, Saint Joseph University (USJ), Lebanon, and a PhD in Edge Artificial Intelligence from the University of Genoa, Italy in 2022.
My expertise spans machine learning, MLOps, and explainable AI. Throughout my career, I have held pivotal roles that blend technical prowess with strategic oversight. At EDT&Partners (AWS Partner), I have been instrumental in guiding companies through the complexities of AI integration, focusing on upskilling teams and enhancing existing infrastructures. As an independent technical AI consultant, I have spearheaded over ten projects, delivering tailored AI solutions that drive operational efficiency and innovation.
Previously, as the AI Lead at Assentify, I led the development of a comprehensive eKYC platform for the InsurTech sector, incorporating advanced features such as ID verification, tampering detection, and dynamic insurance pricing powered by AI. My role also included mentoring sister companies’ AI teams in MLOps environments, ensuring seamless integration and maintenance of their AI products.
During my tenure at MYWAI in Italy, I focused on Edge AI applications, developing predictive maintenance solutions for industrial machinery. These solutions improved maintenance protocols and reduced downtime for clients like MK MBC Russia and Mitsubishi Italy.
In addition to my professional engagements, I am deeply committed to the academic and AI communities. As a previous Adjunct Professor at Saint Joseph University of Beirut, I have taught advanced AI courses, imparting cutting-edge knowledge to the next generation of AI professionals. My research has been published in top-tier journals and presented at leading conferences, highlighting my contributions to the field. I actively engage with the AI community through workshops, podcasts, and speaking engagements with esteemed organizations such as DeepLearning.AI, MDSF, Educative, and Zaka.ai. My work is driven by a commitment to advancing AI technology in innovative and ethical ways.
My passion for AI is rooted in a dedication to advancing technology and its practical applications, ensuring that AI solutions are both innovative and ethically responsible. I continue to explore new frontiers in machine learning, MLOps, and explainable AI, pushing the boundaries of what is possible in this ever-evolving field.
Education
2022
University of Genoa (UniGe), Italy
PhD in Edge Artificial Intelligence
2019
Saint Joseph University – ESIB, Lebanon
Masters in Software Engineering
Languages
- Arabic: Native
- English: Fluent
- French: Fluent
- Italian: Intermediate
Work Experience
2024 - Present
American University of Beirut (AUB)
Lecturer
- Leading AI research and educational initiatives.
- Teaching and mentoring students on cutting-edge AI technologies.
2023 - Present
EDT&Partners, Spain
Senior AI Consultant
- Leading AI research and Generative AI initiatives, while helping companies develop their presence in AI by ideating, planning and developing new features and improving existing systems.
- Helping our strategic partner, AWS, in the development of AWS Lecture.
- Collaborating closely with clients to enhance their AI capabilities and infrastructure.
- Project Example:
- Developing a personalized content delivery system / Proof of Concept (PoC) for Little Thinking Minds (LTM), UAE, an Arabic language teaching platform for native children and beginners, using Python, AWS Bedrock, AWS EC2, and sklearn. The project aimed to enhance user engagement by tailoring educational content to individual interests and reading abilities, and implementing AI-powered features to assess users’ reading proficiency, preferences, and historical usage, providing personalized content recommendations.
2023 - Present
Freelance
Technical AI Consultant
- Providing AI consulting services to various companies already involved in AI and those seeking to enter the AI space, with a focus on product development and team upskilling. Over 10 projects were successfully completed, delivering tailored AI solutions and enhancing client capabilities.
2022 - 2024
Assentify, Lebanon
AI Lead
- Spearheading the development of a comprehensive eKYC platform tailored for the InsurTech sector.
- Leading a talented AI team to deliver key features.
- Implementing robust mechanisms for reading and verifying ID cards and passports.
- Developing advanced algorithms for document tampering detection.
- Creating a reliable facial recognition system and dynamic insurance pricing models.
- Leveraging AI to adjust insurance pricing dynamically based on various factors.
- Ensuring compliance analysis through social media and web scraping.
- Building predictive models to assess health insurance risks accurately.
- Successfully deploying the solution in the international market, handling 20,000 requests per day.
2022 - 2024
Saint Joseph University (USJ), Lebanon
Adjunct Professor
- Teaching advanced AI courses for Masters in Artificial Intelligence students, focusing on practical applications and latest research trends.
2019 - 2022
MYWAI, Italy
AI Research Engineer
- Conducting research and developing cutting-edge applications in the field of Edge Artificial Intelligence.
- Working on several key projects including:
- Predictive Maintenance for Steel Recycling Motors: Designing a system for detecting motor anomalies, using sensor data and Python with Scikit-learn (R&D), translated to C for deployment on microcontroller hardware technology, improving maintenance efficiency for MK MBC Russia for 60 continuous motors.
- Predictive Maintenance for Robotic Arms: Developing a solution for real-time anomaly detection and maintenance notifications via WiFi or BLE5, for Mitsubishi Italy, using Python and Scikit-learn for R&D, and translating the solution to C for deployment on microcontroller hardware, designing and implementing a comprehensive software system for data collection and visualization, significantly enhancing the reliability and efficiency of the robotic arms.
Courses
Online Courses
- Transformers in Computer Vision – Educative.io
- Engineering Performant and Trustworthy AI Solutions – O’Reilly
University Courses
- Machine Learning for Computer Science Students – Delivered in Fall 2023 at Saint Joseph University
- AI in Industry for Masters in AI Students – Delivered in Spring 2024 at Saint Joseph University
- Introduction to Machine Learning (EECE490) – Delivered in Fall 2024 at the American University of Beirut
- Data Centric Python (EECE231) – Delivered in Fall 2024 at the American University of Beirut
Research
2019 - Convolution Neural Networks for Arabic Font Recognition
15th International Conference on Signal-Image Technology & Internet-Based Systems (SITIS)
Developing a smart font recognition system using convolutional neural networks (CNNs) to identify Arabic fonts from images. A new dataset of 2500 images covering 50 Arabic fonts was created, and various deep learning models like AlexNet and ResNet were trained. The best model used a sliding window technique to classify full text documents, significantly aiding designers in font identification.
2021 - Maritime Localization System Based on IoT
28th IEEE International Conference on Electronics, Circuits, and Systems (ICECS)
Presenting a low-cost, efficient localization system for maritime applications, utilizing Bluetooth 5.1. The system, capable of indoor and outdoor localization, demonstrated accuracy of 1.1 meters indoors and 7.3 meters outdoors. It visualizes targets on a dynamic web map, enhancing safety and operational efficiency in maritime environments.
2021 - Experimental Assessment of Moving Targets Localization Performance Based on Angle of Arrival and RSSI
AISEM Annual Conference on Sensors and Microsystems
Assessing a localization system using Bluetooth technology, combining Angle of Arrival (AoA) and Received Signal Strength Indicator (RSSI) to track moving targets. The study highlighted the system’s potential for real-time indoor localization, providing a cost-effective solution for various applications.
2022 - A Convolutional Neural Network-Based Method for Discriminating Shadowed Targets in Frequency-Modulated Continuous-Wave Radar Systems
Sensors 22 (3), 1048 Journal
Addressing the challenge of target discrimination in FMCW radar systems caused by radar shadow effects. A novel CNN method was developed to analyze spectrograms from radar signals, achieving a 92% accuracy in distinguishing shadowed targets, thereby improving the reliability of low-cost radar systems.
2023 - On Edge Human Action Recognition Using Radar-Based Sensing and Deep Learning
IEEE Transactions on Industrial Informatics Journal
Developing a radar-based human action recognition system deployed on edge devices. Using deep neural networks to analyze Range-Doppler maps from FMCW radar, the system achieved a 93.2% accuracy in recognizing five human actions and a 96.8% accuracy in fall detection, proving its utility in indoor safety applications.
2023 - Localization Techniques in Cellular Communications
Exploring the use of AoA and RSSI localization algorithms in real-world scenarios, focusing on applications in military, medical, and industrial contexts. Two systems tested in various environments were presented, demonstrating the viability of these low-cost solutions for short-range localization up to 150 meters.
For more detailed information, please refer to my Google Scholar Profile.