Rohan Mahfouzi

A solution that upgrades any touch screen with a touchless interaction mode that does not interfere with the existing touch-based user experience. Users can use hand gestures to interact with computers without touching the screen. The solution is easy to integrate, accurate, and reliable. We used machine vision and artificial intelligence techniques to enable touchless interaction. The project's user interface is developed using Qt technology, while the core functionality and learning algorithms were developed in C++ to ensure the best performance.

A C++-based implementation of AES and DES encryption protocols was used in this project and the aim was to measure the encryption and decryption performance of the ARM processor used in NVIDIA® Jetson kits at a certain frequency.
The ultimate goal was to research the relation between the number of rounds of encryption and different key sizes versus the timing overhead of running the encryption and decryption.

Real-time software for monitoring the status of the boat, where the interface is designed and developed using Qt/QML and is developed in such a way that it has more than one graphic design. The software itself is robust as it must function for a long time in marine applications.

Software that connects to the centrifugal apparatus and captures real-time core images from a high FPS camera.

An analyzer module calculates the saturated and unsaturated volumes in the next step by employing image processing techniques. Then, a simulation core uses this data for history matching to find relative permeability parameters. The simulation code of the software is developed from scratch and integrated as a standalone module. A YouTube link to the demo of the software is provided.

I was the senior developer and team lead in this project leading a team of 3-4 software engineers. I had a broad overview of the architecture of the software while also helping the implementation team. I was also the contact point for the correspondent petrochemical engineer from the client.

An application with multiple plots and histograms to visualize data from medical equipment, where I implemented the UI with Qt and connected it to the back-end codes and the technologies used were Qt, QCustomPlot, and C++.

A smart sensor developed for environmental monitoring and security automation, designed to detect vape, THC, and air quality. The device incorporates edge computing capabilities to run advanced sound classification models, enabling it to recognize specific spoken keywords and accurately classify critical acoustic events such as gunshots, glass breaking, and aggression. The device also uses radar to detect occupancy and loitering. Built on a custom embedded firmware architecture, the system utilizes integrated microphone arrays and sensor modules to deliver real-time, low-latency threat detection and automated alerts.