In 2024, the RubiQ team and I won the Pasqal-CMC hackathon, where we tackled the complex challenge of optimizing antenna placement across multiple cities to maximize 5G coverage while avoiding overlap. Our solution used an innovative combination of simulated annealing and Pasqal's Pulser library to leverage Neutral Atom Quantum Computing (NAQC) in solving a Maximum Weighted Independent Set (MWIS) problem for optimal coverage distribution.

Our primary approach combined classical methods with quantum resources: simulated annealing generated grid layouts, while NAQC determined MWIS solutions efficiently, capitalizing on Rydberg atom interactions. This hybrid strategy enabled us to maximize coverage without antenna overlap.

We also explored a second approach using the Variational Quantum Annealing Algorithm (VQAA), optimizing atom detunings and pulse parameters. While limited by current simulation capabilities, this method shows promise for future applications with larger qubit arrays, as it offers a fine-grained optimization potential that could be scaled for complex network scenarios.

During the 7th Montreal Photonics Networking Event, I presented my research on the inverse design of nonlinear materials.

I developed and released an open-source library implementing a genetic algorithm to optimize second-order nonlinear crystals for enhanced entangled photon generation through Spontaneous Parametric Down-Conversion (SPDC).

This work advances quantum light source technology by refining material characteristics crucial for high-precision applications in quantum imaging and metrology.

For more information, see the GitHub link above and the brief article!

From 2022 to 2024, I spent two years working at the Laboratoire des Fibres Optiques under the supervision of Stéphane Virally, gaining invaluable practical and theoretical knowledge.

My work focused on the experimental realization of "band-conditioned states" — exotic quantum light states with significant potential in metrology. This process involves generating entangled photon pairs, measuring one branch in phase space, and conditioning the measurement on the intensity of the other branch. This post-selection enables us to obtain Wigner functions with distinct negative regions (in blue on the image above), characteristic of quantum properties in this quasi-distribution.

The experimental realization of such states is challenging, but have made significant progress!

In this project, I focused on developing an ultrafast Erbium-doped fiber laser system specifically designed for quantum optics applications. The main goal was to create an ultralow-noise, high peak power pulse train capable of efficiently driving nonlinear frequency mixing processes.

To achieve this, we built a passive mode-locked master oscillator, further amplified by an Erbium-doped fiber amplifier (EDFA).

After several optimization steps, we achieved pulse durations around 139 fs FWHM with an energy of approximately 4 nJ. We also characterized the laser's intensity noise, obtaining an integrated RIN of 0.081%.

During the 2022-2023 academic year, I led a cooperative project with Camsekin in Kinshasa to tackle the challenges of an unreliable electrical grid in the Democratic Republic of Congo. Our objective was to ensure continuous operation of a solar-powered, refrigerated medical storage facility by designing an advanced energy monitoring system.

This initiative, developed in partnership with Camsekin and the École Polytechnique de Bruxelles, focused on enhancing battery resilience through precise monitoring and predictive analysis. The system, adaptable to rural setups, tracks battery life with a Kalman filter-based model and visualizes real-time data remotely, supporting critical infrastructure in resource-limited environments.

Using an Arduino DUE microcontroller, we implemented scripts in Python to monitor solar panel output, comparing actual performance with local weather data predictions. This pilot project is poised to expand across rural medical facilities, safeguarding essential cold-chain storage systems.

For more details, see the GitHub link above!

This project consists in developing a C program to solve the two-dimensional Poisson equation numerically. The focus was on modeling different shapes of membranes and solving the equation using an iterative Multi-Grid method. A key aspect of my project was optimizing the solution process with a relaxation parameter and enhancing the solver's performance by integrating a Multi-Grid preconditioner alongside the PRIMME solver.

The multi-grid method is an efficient numerical approach for solving differential equations, leveraging multiple levels of grid resolution. The core idea behind the two-grid algorithm is to perform most computational steps inexpensively, with the exception of solving with the UMFPACK direct solver on the coarse grid. The multi-grid method enhances this by iteratively applying pre-smoothing and post-smoothing on progressively coarser grids. Each new coarse grid doubles the discretization step of the old grid until a grid is coarse enough for the problem to be solved directly, significantly reducing computational costs.

For more information, see the GitHub link above!

My interest has always been in figuring out how things around me work. This curiosity led me down an engineering path in physics, where every answer brought more questions. Along the way, I developed a deep interest in quantum mechanics, the cornerstone of my understanding of the universe. However, my penchant for tangible evidence made me more of an experimentalist than a theorist.

Enrolling in Brandenberger Robert's course on quantum field theory at McGill University was an opportunity for me in my quest for knowledge. The course introduced me to the nuances of a framework that, despite its relative youth and certain vulnerabilities, forms the foundation of the powerful Standard Model of particle physics and holds out the possibility of a unified theory.

In this context, I wrote an essay on Hawking radiation, a theoretical prediction where black holes emit thermal black body radiation. Following Stephen Hawking's seminal paper, I used a simplified model involving massless scalar fields for a Schwarzschild black hole to deduce the black hole's blackbody radiation.

For additional information, see the essay linked below!

During my last year of high school, I challenged myself to program a video game within a few months, aiming to publish it on the App Store. After dedicated effort and substantial learning, I successfully released a retro arcade game named KURUMA (meaning "car" in Japanese).

It was a lot of work, but I loved the process and learned a tremendous amount along the way. From coding and troubleshooting to designing gameplay mechanics, each step deepened my understanding of software development.

During this project, I developed a ray-tracing software to analyze electromagnetic wave propagation, simulating power reception from WiFi access points and mapping 5G indoor small cell base station coverage areas, including received bit rate based on receiver location.

To ensure efficiency, we used key assumptions: lossless half-wavelength dipole antennas for 5G, uniform transmitter/receiver height, horizontal wave propagation, and a far-field approximation, focusing only on wall reflections and transmissions up to three times while ignoring diffraction and internal obstacles.

The C++ software featured three modes: visualizing received rays and power, mapping coverage areas with variable transmitter placements, and optimizing transmitter count and location using a genetic algorithm inspired by C. Ting’s work. This innovative approach used chromosome replication to streamline positioning, maximizing coverage without predefined source numbers.

For more details, see the GitHub link above!

This Tower Defense game, developed in Java, challenges players to survive numerous waves of diverse monsters by strategically placing and upgrading various weapons across multiple maps. Each monster possesses unique speed and health attributes, adding complexity to the defense strategy. Players earn in-game currency by defeating monsters, which can be used to upgrade weapons and enhance their defense capabilities. A standout feature of the game is the ability for players to design their own maps, offering a personalized gaming experience and endless replayability.

For more information, see the GitHub link above!