Profile Photograph
Clément Rouvroy

I am a Computer Science Master's student at ENS PSL (Paris), specializing in the intersection of AI and data systems.
My research interests include Databases, Optimization, AI, ML4DB, and DB4AI.

Curriculum Vitae

My full Curriculum Vitae is available for download here.

Here are my primary areas of interest and the related courses I completed. All courses were completed with the highest distinction (mention très bien), except for those in italics which are part of my current curriculum.

My academic background has provided me with both the practical skills for high-efficiency coding (e.g., building a compiler, a Git clone from scratch, and many AI projects) and the deep theoretical knowledge in areas like Linear Algebra, Complexity Theory, and Database Theory.

Artificial Intelligence
Data
Applied Computer Science
  • OS (Timothy Bourke)
  • Numerical System
  • Compilation (Jean-Christophe Filliâtre)
Formal & Algorithmic

Projects

Reinforcement Learning for Autonomous Cars
AI

A simple car simulation, with controls comparable to 2D Trackmania, learns to follow complex routes using reinforcement learning.

Hypothetical Index for Column-Oriented Databases
Database
Hypothetical Index Overview

This project provides a lightweight "what-if" index benefit estimator designed specifically for column-oriented storage, using new heuristics tailored to the columnar architecture. Detailed explanations and results are available in the internship report.

Diffusion Models for Galaxy Generation
AI
Galaxy generation with diffusion models

Implemented diffusion models from scratch as part of a university course, applying them to the task of procedurally generating images of galaxies.

GNN for Fake News Detection
AI
GNN for Fake News Detection

Implemented a fake news detector for Twitter using Graph Neural Networks (GNNs), based on the approach from the paper "Fake News Detection on Social Media using Geometric Deep Learning".

RISC CPU in Netlist with Custom Assembly
Theory
CPUlm Netlist diagram

CPUlm is a RISC processor designed and implemented in Netlist. It supports arithmetic-logic instructions, conditional jumps, and memory access. Full documentation for the instruction set and custom assembly language is available.

PureScript Compiler
Theory

Built a full compiler for a subset of PureScript. It includes a lexer, parser, a type-checker with inference (handling type classes beyond Algorithm W), and an x86 assembly code generator.

Git Clone in C++

Developed GitAne, a highly customizable, plugin-based Git clone in C++. The project includes a functional, authenticated server for hosting repositories.

Research Experience

For a detailed list of my publications, please see my research page.

My research focuses yet on optimization for data bases, though I am now focusing on AI and its intersection with data bases.

NTU

Research Internship at NTU Singapore

Time Period: February - July 2025
Location: MLxDB of NTU (Nanyang, Singapore)

Under the supervision of Gao Cong and Jiachen Shi, my work focuses on developing new index recommendation opportunities for modern database systems.

INRIA

Research Internship at INRIA

Time Period: June - August 2024
Location: BOREAL Team of INRIA (Montpellier, France)

I worked with Nofar Carmeli and David Carral on the enumeration of acyclic conjunctive queries with self-joins. A presentation of this work is available here.

ENS-PSL

Research Project at ENS-PSL

Time Period: September 2024 - January 2025
Location: Valda Team of ENS-PSL, Inria Paris and CNRS (Paris 05, France)

I continued my summer research on query enumeration as part of the M1 curriculum, supervised by Luc Segoufin within the Valda team at ENS.

Contact

You can reach me via email at [given name].[family name]@ens.psl.eu.