About

I study Data Science & AI at ESIEE Paris and Physics at Sorbonne University (parallel track).
Co-founder of Reaply, an AI review-intelligence platform for multi-site brands.
Focus: applied ML, careful evaluation, and robust production systems.

Availability: May–Aug 2026, full-time research internship.
Interests: computer vision, discrete geometry & topology, robust learning, fairness, ML systems.

Now

Ongoing Research

Research program — Extending the Morse Frames framework
Advisors: Prof. Laurent Najman, Prof. Gilles Bertrand

Topic. Morse sequences and image segmentation via discrete Morse theory.
Goals.

• Build discrete Morse complexes on 2D/3D images
• Define topological criteria for region merge/split
• Benchmark against watershed, graph cuts, and modern CNN baselines
• Integrate Morse frames into a reproducible CV pipeline

Methods. Discrete Morse complexes, topological persistence criteria, structured pipelines in Python; unit-tested operators; experiment tracking.
Evaluation plan. Standard segmentation metrics (IoU, boundary F1), ablations on merge/split rules, stress tests on noise and resolution.

Selected Work

Reaply — Review-intelligence platform (co-founder)

Production ML/NLP stack for multi-site brands:

• Ingestion and normalization of multi-source reviews at scale
• Sentiment and topic models for trend and drift detection
• Assisted, policy-aware reply generation
• Data contracts, monitoring, and CI for models and pipelines

Role: ML lead and systems engineering. From prototype to production with reliability and evaluation discipline.

Teaching and Community

• First-year tutoring (university): mathematics, physics, Java basics, electronics.
  Teach and assist during sessions, answer questions, provide concise written feedback after class.
• Secretary-General, Jeunes Français de l’Étranger (JFDE): representing young French expatriates to the French government.

Open to

Research or engineering internships where applied ML meets impact.
Topics welcome: computer vision, discrete/geometry-aware learning, robustness, fairness, ML systems.