Curriculum vitae · September 2026

Experience,
education & work.

Machine-learning researcher and engineer with a foundation in statistics, deep learning, and research software.

Contact
01

Profile

Graduate of École polytechnique and ENSAE Paris—both member schools of Institut Polytechnique de Paris (IP Paris)—in data science and statistics, available from mid-October 2026. I am interested in understanding model behavior and in building reliable experimental and software tools for machine-learning research.

02

Experience

KU Leuven · LIIR

Research intern · Leuven, Belgium

KU Leuven · 59th worldwide · QS World University Rankings 2027 ↗

  • Studied cross-lingual access failures after knowledge editing in language models, through causal interventions on model representations.
  • Built reproducible PyTorch evaluation pipelines on GPU: multilingual medical benchmarks, ablations, causal controls, and checkpoint audits.
  • Work conducted towards an ICLR 2027 submission (September 2026).

ENSAE Junior Entreprise

Data science consultant · Paris, France

  • Developed TrustAgent, an LLM-based document analysis and decision-support system, in Python, FastAPI and LangGraph.
  • Owned the comparison-and-decision component: multi-model orchestration and citation traceability.
  • Automated tests, reproducible benchmarks, technical documentation, and Docker deployment.

Groupe BPCE

Research data science intern · Paris, France

  • Reviewed and implemented research methods in Positive–Unlabeled learning.
  • Built an internal Python package for scoring and classification.
  • Presented the results to the group’s scientific committee of 100 attendees.
03

Education

ENSAE Paris · IP Paris

Engineering degree (Diplôme d'ingénieur) · Specialization in Data Science, Statistics & Machine Learning (DSSA) · Direct entry into the second year after ENSAI

IP Paris · Joint 43rd worldwide · QS World University Rankings 2027 ↗

Advanced machine learning, deep learning, NLP, generative models, Bayesian statistics, econometrics, time series, optimal transport, deployment of data science projects.

ENSAI

Engineering degree · Statistics & Computer Science · Top 5%

Statistical learning, inferential statistics, measure theory, algorithms, and software engineering.

Lycée Descartes · Tours

Preparatory classes · MPSI / MP

04

Selected projects

Study of the CANDI diffusion model

Generative-models project

  • Studied noise schedules and crystallization in continuous-and-discrete diffusion.
  • Adapted training to WikiText-2 on an NVIDIA A100.
  • Compared WideResNet and CNN capacity under iterative attacks on CIFAR-10 and GTSRB.
  • Studied stochastic training as a regularizer for the accuracy–robustness trade-off.

Surrogate Modeling Toolbox ↗

Open-source contributor

  • Implemented Gaussian-process variance re-interpolation methods for Bayesian optimization.
  • Contributed optimization code, unit tests, and documentation.