About Me

In August 2021 I joined the Programming Group at the University of St. Gallen, Switzerland as a PhD student where I'm supervised by Guido Salvaneschi.

Between 2017 and July 2021 I was an R&D software developer at Agilent Technologies (formerly Genohm) in Lausanne, Switzerland.

In June 2017 I obtained a master's degree in computer science from the École polytechnique fédérale de Lausanne (EPFL) in the LAMP lab. In my thesis I simplified conceptually a polymorphic safe-and-precise effect system.

In 2015 I obtained a bachelor's degree in computer science and a minor in mathematics from the American University of Beirut (AUB). My bachelor's graduation project was to design and implement a quantum functional language.

I am an avid reader, my Goodreads profile lists all my non-technical interests (those outside of CS and math). I speak/read fluently three (natural) languages: English, French, and Arabic.

In my spare time, when I'm not reading or programming, I play Go, snooker, and the mandolin.

Academic Interests

• Formal study of Programming Languages with a focus on safe type systems,
• Mathematical Logic and its relation to programming,
• and Quantum Computation from the perspective of computability: here's an introductory article about Quantum Computation ^[PDF].

Technical Knowledge and Setup

I am proficient in many programming paradigms; functional, imperative, and object-oriented, with preference to functional languages. I mainly use Python for small applications, Bash for write-once-execute-once code, C and D for any heavy lifting, Octave for quick plotting and mathematical computations and Haskell for any type-theoretic experiments and programs that must be correct. I am proficient in these programming/scripting/markup languages:

• C, D, Haskell, Scheme, Java, Scala
• Python, Bash, Javascript, PHP
• HTML, CSS, and LaTeX

My primary editor is Vim.

I live in the command line most of the time.

Publications

• George Zakhour, Pascal Weisenburger, and Guido Salvaneschi. 2023. Type-Safe Dynamic Placement with First-Class Placed Values. Proceedings of the ACM Programming Languages 7, OOPSLA2, Article 297 (October 2023), 29 pages. https://doi.org/10.1145/3622873 [author version]
• George Zakhour, Pascal Weisenburger, and Guido Salvaneschi. 2023. Type-Checking CRDT Convergence. Proceedings of the ACM on Programming Languages 7, PLDI, Article 162, June 17–21, 2023, Orlando, Florida, United States. https://doi.org/10.1145/3591276 [author version]

Projects

• System F-CA My masters thesis that builds on the work of Fengyun Liu where we design a precise-and-safe polymorphic effect system using capabilities using one kind of functions (kind here does not refer to the standard type-theoretic definition).
• Braqet and the MG-QTM Simulator Braqet is the functional quantum programming language I developed for my bachelor's degree. Its type system adopts the linear types introduced by Selinger and Valiron (2006) with the exception that control in Braqet is purely quantum. The MG-QTM Simulator simulates MG-Quantum Turing Machines and visualises them. MG-QTMs are more general SR-QTMs ^[PDF]. For the visualization of the graph I used a Forced-Directed Graph Drawing Algorithm ^[PDF].
• Damncheck A property-based testing library for D. [Source Code]
• Simulation of a Quantum Particle in a Box [Source Code] [PDF Report]
• e Project I found the first 5,087,137 digits of e in ~1hr29min. In 2019 I wrote a parallel version that descreases the runtime linearly with the number of processes and ended up finding the same digits in 27min5sec with 8 threads. [Talk]

What is this graph above?

This graph represents a 69-state machine, in fact, a quantum Turing machine ^[PDF]. It's a machine that abides by the laws of quantum mechanics. This machine can be thought of performing the following algorithm on each superposition |k,x⟩ in a 6-dimensional Hilbert space

x ∈ ℕ8
k ∈ ℕ8
while |x-k| < 4:
    x := x + 1   mod 8

This machine is significant to me because I devised it to exhibit conditional looping (as opposed to a statically-bounded loop) which I could not find in the literature. In the process I discovered that the fixed-point of a sequence of compositions of unitary functions may not be itself unitary which raises the question of computing the inverse of such a function.