Wael Fatnassi, Ph.D. Email & Phone Number

New York, United States

• Led the development and management of the "NOVOS Life" AI-Chatbot, focused on healthcare-related topics.
• Collaborated with R&D leadership, engineers, and app developers to optimize and enhance the chatbot’s performance through rigorous testing, troubleshooting, and deployment.
• Implemented new techniques and conducted stress testing to ensure the chatbot could support millions of users.
• Improved user engagement during the pre-release phase, resulting in higher-than-expected user interaction and retention.

Irvine, California, United States

• Developed an algorithm for assessing the safety of neural networks, achieving response times 400% faster than the AlphaCrown algorithm, with execution times measured in millionths of seconds.
• Implemented an advanced solver using neural networks and Bernstein polynomials in the context of self-driving cars, surpassing the performance of established solvers like Linear Programming and Microsoft's Z3.
• Improved safety of autonomous plane landings by engineering a controller program using machine learning and airplane camera vision, and conducted simulations through Microsoft’s XPlane to validate its effectiveness.
• Created path-planning and implemented PID controller for an autonomous system (Raspberry PiCar).

United States

• Developed a chest X-ray image classifier for detecting COVID-19 using transfer learning.
• Designed text generation using Markov Chains and deciphered text through character-level RNNs.
• Single-handedly built a language model using a recurrent neural network from scratch.
• Engineered an innovative NLP model leveraging transfer learning to predict emojis, facilitating more expressive and contextually accurate communication.
• Developed an object detection system for thermal vision using PyTorch and YOLOv5, enhancing night-time surveillance capabilities for various applications.

United States

• Led 2 classes per week to guide 100 undergraduate engineering students through the fundamentals of perception, planning, and control of autonomous systems, as well as the Robot Operating System (ROS).
• Designed course materials to teach data integration & autonomous system algorithms to 100+ Master’s students.
• Regularly presented research and product demos to audiences of 500+ at HSCC, Toyota Institute, and Qualcomm.

Calabasas, California, United States

• AI-Driven Strategic Analysis of CSGO Gameplay Using GPT-2 Models:
• Developed machine learning models to analyze complex systems involving interactions between humans and autonomous agents during critical missions.
• Created a GPT2-based machine learning model to explore dynamics between human players and autonomous agents in Counter-Strike Global Offensive (CSGO) competitive matches.
• Build a data pipeline to preprocess and structure large E-sport datasets from professional CSGO matches for efficient model training.
• Demonstrated the model's ability to forecast pivotal gameplay events, enhancing understanding of team performances and strategic decision-making.
• Successfully identified unexpended surprising events in games, facilitating insights into unconventional tactics that influence match outcomes.Dynamic ROS2 and Autoware Integration for Autonmous Vehicles:

Idaho, United States

• Trained Deep Neural Networks such as Autoencoder for 1) Symbol Detection and Channel Estimation in Optical Wireless Communications (OWCs) systems and 2) Partial Interference Cancellation in Uplink Cellular Networks.
• The proposed method achieved more than a 90% success rate in detecting the symbols and reducing user interference.

Idaho Falls, Idaho, USA

• Presented the physical layer security aspects of 5G networks to INL's leadership team.
• Led risk-benefit analyses and ROI assessments to guide INL in adopting state-of-the-art 5G enabling technologies such as millimeter waves (mmWaves), mobile edge computing (MEC), and hybrid beamforming.
• Assessed the viability of existing experiments and testbeds in supporting 5G enabling technologies.

Moscow, Idaho, United States

• Studied the reliability of the Smart Metering System (SMS) and provided the quantifiable impact of path loss, fading, shadowing, and co-channel interference on SMS.
• Modeled Multiple Access Channels (MAC) for 5G millimeter wave frequencies for SMS reliability analysis.

Leipzig Area, Germany

• Designed a new algorithm for Siemens to reduce interference in their transmit-only smart meters by programming Panstamps and USRP using languages C and Linux.
• The algorithm achieved more than a 93% success rate in reducing interference.

Tunis, Tunisia

• Provided critical research to support the company’s 2G to 3G migration plans through a detailed report on the Universal Mobile Telecommunications System (UMTS) network architectures.
• Developed an application that determines the optimum diameter of the cellular zone considering the estimated number of users and Network flows using languages C# and Ubuntu.

Tunis, Tunisia

• Provided intensive testing and documentation for new equipment within the cellular network division to ensure reliability and safety before roll-out.

Doctor Of Philosophy - Phd, Electrical, Electronics And Communications Engineering, 4.0/4.0

Advisor: Yasser Shoukry Relevant Courses: Algorithms and Data Structure, Machine Learning System Design, Network Science, Optimization.

Master Of Science - Ms, Electrical Engineering, 3.9/4.0

Advisor: Zouheir Rezki Thesis Title: Learning-Based Communication Systems. Relevant Courses: Deep Learning and Spiking Neural Networks.

Bachelor Of Science - Bs, Telecommunications Engineering, 17.5/20.0

Higher School of Communications of Tunis Thesis Title: MIMO Systems for Increasing the Reliability of the Wireless Communications in.

Mathematics And Computer Science

Preparatory School For Engineering Studies of Tunis (ie Pre-Engineering Studies) Special Recognition: Ranked in the top 3% of a Tunisian.