cv.yml

- title: General Information
  type: map
  contents:
    - name: Full Name
      value: Louis Berthier
    - name: Position
      value: PhD Candidate, Applied Mathematics & Machine Learning
    - name: Affiliation
      value: <a href="https://cmap.ip-paris.fr/">CMAP</a>, École Polytechnique (IP Paris), in collaboration with <a href="https://www.michelin.com/">Michelin</a>
    - name: Email
      value: <a href="mailto:louis.berthier@polytechnique.edu">louis.berthier@polytechnique.edu</a>
    - name: Website
      value: <a href="https://louistier.github.io/">louistier.github.io</a>
    - name: Languages
      value: French (native), English (fluent), Spanish (professional)

- title: Education
  type: time_table
  contents:
    - title: PhD Candidate in Applied Mathematics and Machine Learning
      institution: <a href="https://cmap.ip-paris.fr/">École Polytechnique (CMAP)</a>, Paris, France
      year: 2024 - now
      description:
        - "Conformal and adaptive soft sensors for real-time quality prediction in industrial batch processes."
        - Under the supervision of <a href="https://www.linkedin.com/in/eric-moulines-61512860/">Pr. Eric Moulines</a>, <a href="https://www.linkedin.com/in/ahmed-shokry-aa705075/">Dr. Ahmed Shokry</a>, <a href="https://www.linkedin.com/in/desroziers/">Dr. Sylvain Desroziers</a>, <a href="https://www.linkedin.com/in/moreaudmaxime/">Dr. Maxime Moreaud</a>, and <a href="https://www.linkedin.com/in/guillaume-ramelet-7696081/">Guillaume Ramelet</a>.
        - "Adaptive Soft Sensing, Conformal Prediction, Explainability, Self-Organizing Maps, Online Learning, Dimensionality Reduction."
    - title: MSc Advanced Computing in Machine Learning
      institution: <a href="https://www.imperial.ac.uk/">Imperial College London</a>, London, UK
      year: 2022 - 2023
      description:
        - "Diploma with merit (2:1), and thesis with distinction (1st)."
        - Research internship at the <a href="https://www.imperial.ac.uk/adaptive-intelligent-robotics/">Adaptive & Intelligent Robotics Lab (AIRL)</a>, under <a href="https://www.imperial.ac.uk/people/a.cully">Pr. Antoine Cully</a>.
        - "Machine & Deep Learning, Reinforcement Learning, Computer Vision, Natural Language Processing, Mathematics for Machine Learning, Computational Finance, Robot Learning."
    - title: "Diplôme d'Ingénieur, MSc Artificial Intelligence & Data Science"
      institution: <a href="https://www.imt-mines-ales.fr/">IMT Mines Alès</a>, Alès, France
      year: 2020 - 2023
      description:
        - "First class honours with jury's congratulations. Ranked 2nd in CS & AI department, top 5% of the class."
        - Research internship at the <a href="https://cerco.cnrs.fr/">CNRS (CerCo)</a>, under <a href="https://www.linkedin.com/in/ludovic-gardy/">Dr. Ludovic Gardy</a>, <a href="https://www.linkedin.com/in/christophe-hurter-2514393/">Pr. Christophe Hurter</a>, and <a href="https://www.linkedin.com/in/emmanuel-barbeau-153b151b9/">Pr. Emmanuel Barbeau</a>.
        - "Advanced Probabilities & Statistics, Operational Research, Algorithms & Complexity, Machine & Deep Learning, Software Engineering."

- title: Experience
  type: time_table
  contents:
    - title: Computer Vision Scientist
      institution: <a href="https://www.augmodo.com/">Augmodo</a> (part-time alongside PhD)
      year: 2025 - now
      maindescription: Uncertainty-aware computer vision for open-world retail deployments, processing millions of images daily.
      description:
        - Working alongside <a href="https://www.sachahu.com/">Sacha Hu</a> (Senior CV Engineer), <a href="https://www.linkedin.com/in/ashleyanapier/">Ashley Napier</a> (CV Group Lead), and <a href="https://www.linkedin.com/in/shubhamwagh/">Shubham Wagh</a> (Senior CV Engineer).
        - "Building uncertainty-aware classifiers for product identification with test-time augmentation, reducing manual QC workload by selectively accepting high-confidence predictions."
        - "Developing an RGB+depth pipeline to flag empty shelf areas, focusing human review on genuinely uncertain cases."
        - "Designing low-cost dataset relabeling strategies to update legacy bounding box annotations at scale using improved detectors."
        - "Fine-tuning object detectors (YOLO and the transformer-based RF-DETR) for empty shelf-area detection, leveraging hard-negative examples to push detection performance; handling training and fine-tuning on cloud GPUs."
    - title: Research Engineer (CIFRE PhD)
      institution: <a href="https://www.michelin.com/">Michelin</a>, Clermont-Ferrand, France
      year: 2024 - now
      maindescription: Adaptive soft sensing and conformal prediction for real-time quality monitoring in rubber manufacturing.
      description:
        - Supervised by <a href="https://www.linkedin.com/in/desroziers/">Dr. Sylvain Desroziers</a>, <a href="https://www.linkedin.com/in/moreaudmaxime/">Dr. Maxime Moreaud</a>, and <a href="https://www.linkedin.com/in/guillaume-ramelet-7696081/">Guillaume Ramelet</a>.
        - Developed <a href="https://github.com/michelin/TorchSOM">TorchSOM</a>, the reference PyTorch library for Self-Organizing Maps, released under Apache 2.0 via Michelin's open-source organization (74-98% faster than MiniSom on CPU, GPU-accelerated, scikit-learn API, 90% test coverage).
        - "Designed and benchmarked five adaptive soft sensing strategies for real-time quality monitoring, validated on 35,125 production batches, cutting prediction error by around 59% short-term and over 90% under year-long drift."
        - "Developed Self-Organized CP and Masked CP, two conformal prediction frameworks for locally adaptive and mask-valid sequential uncertainty quantification in industrial settings."
    - title: Teaching Assistant (Machine Learning & Data Science)
      institution: <a href="https://www.ensta-paris.fr/">ENSTA Paris</a>
      year: 2024 - 2025
      maindescription: Led machine learning lab sessions for first-year engineering students, with 20 to 25 students per session.
      description:
        - Taught the fundamentals of machine learning, from data preprocessing and feature engineering to model training and evaluation, under <a href="https://perso.ensta-paris.fr/~philippe.xu/">Pr. Philippe Xu</a>.
        - "Guided students through their first hands-on experience building and validating predictive models."
    - title: Research Assistant (ML & Optimization)
      institution: <a href="https://www.imperial.ac.uk/adaptive-intelligent-robotics/">Adaptive & Intelligent Robotics Lab (AIRL)</a>, Imperial College London
      year: 2023
      maindescription: Sample-efficient Quality-Diversity optimization using surrogate models for robotic control in uncertain environments.
      description:
        - Supervised by <a href="https://www.imperial.ac.uk/people/a.cully">Pr. Antoine Cully</a>.
        - "Tackled the data-hunger of MAP-Elites by integrating probabilistic surrogate models, achieving significant gains in sample efficiency on a robotic arm control task."
        - "Benchmarked eight algorithmic variants combining different surrogate architectures and acquisition strategies in non-deterministic environments."
    - title: Research Intern (ML & Neuroscience)
      institution: <a href="https://cerco.cnrs.fr/">CNRS (CerCo)</a>, Toulouse, France
      year: 2022
      maindescription: Deep learning for detecting fast ripples, pathological biomarkers for epileptic tissue, in intracranial EEG signals.
      description:
        - Supervised by <a href="https://www.linkedin.com/in/ludovic-gardy/">Dr. Ludovic Gardy</a>, <a href="https://www.linkedin.com/in/christophe-hurter-2514393/">Pr. Christophe Hurter</a>, and <a href="https://www.linkedin.com/in/emmanuel-barbeau-153b151b9/">Pr. Emmanuel Barbeau</a>.
        - "Built a CNN pipeline converting raw EEG signals into time-frequency scalograms via continuous wavelet transform for automated fast ripple classification."
        - "Applied Grad-CAM to provide clinicians with visual explanations of which signal features drive each prediction."
        - "Accelerated the processing pipeline by 15% through Cython optimization; improved classification accuracy by 5% via hyperparameter tuning."
    - title: Research Assistant (ML & Computer Vision)
      institution: <a href="https://dhm.euromov.eu/">EuroMov Digital Health in Motion (DHM)</a>, Montpellier, France
      year: 2022
      maindescription: Anisotropic image filtering for sub-pixel accurate line feature detection, published at IEEE ICASSP 2023.
      description:
        - Supervised by <a href="https://cv.hal.science/baptiste-magnier">Pr. Baptiste Magnier</a> and <a href="https://www.researchgate.net/scientific-contributions/Binbin-Xu-2190252279">Pr. Binbin Xu</a>.
        - "Designed the 2D Semi Bi-Gaussian (2DSBG) filter, an asymmetric convolution kernel that enhances elongated structures while suppressing isotropic noise."
        - "Benchmarked against state-of-the-art edge detection models (DexiNed, RCF, Laplacian-of-Gaussian) on synthetic and real-world datasets."

- title: Open Source
  type: time_table
  contents:
    - title: <a href="https://github.com/michelin/TorchSOM">TorchSOM</a>
      year: 2024 - now
      maindescription: Reference PyTorch library for Self-Organizing Maps, released under Apache 2.0 via Michelin's open-source organization. 74-98% faster than MiniSom on CPU, GPU-accelerated, scikit-learn API, 90% test coverage.

- title: Honors and Awards
  type: time_table
  contents:
    - title: <a href="https://www.youtube.com/watch?v=VjbWHeqfhMk">MT180 Public Award (€100)</a>
      institution: Institut Polytechnique de Paris
      year: 2025
      maindescription: Won the audience vote for best 3-minute PhD presentation at Institut Polytechnique de Paris, selected to represent IP Paris at the national finals.
    - title: Best PhD Poster Award (€500)
      institution: Institut Polytechnique de Paris
      year: 2024
      maindescription: Best poster in the Mathematics category at the IP Paris Welcome PhD Day, for work on explainable quality prediction with uncertainty quantification.
    - title: Winner, Manufacturing IT Hackathon AI'NNOV (€500)
      institution: Michelin
      year: 2024
      maindescription: First place at Michelin's internal AI hackathon for manufacturing innovation.
    - title: International Mobility Scholarship (€4000)
      institution: Région Occitanie
      year: 2022
      maindescription: Scholarship supporting international academic mobility.
    - title: Alumni Scholarship (€1300)
      institution: Mines Alès Alumni
      year: 2022
      maindescription: Merit-based scholarship from the Mines Alès alumni association.
    - title: Excellence Scholarship (€4000)
      institution: IMT Mines Alès
      year: 2022
      maindescription: Academic excellence scholarship.
    - title: Creativity Laureate, Communication & Media (€80)
      institution: Cora Hypermarket
      year: 2021
      maindescription: Creativity laureate for a Communication & Media competition.

- title: Skills
  type: list_groups
  contents:
    - category: Programming & Scripting (Expert)
      icon: fa-solid fa-code
      items:
        - name: Python, Cython
        - name: SQL, NoSQL
        - name: LaTeX, Markdown
        - name: Bash, Shell
    - category: ML/DL Frameworks (Master)
      icon: fa-solid fa-gears
      items:
        - name: PyTorch, JAX, TensorFlow
        - name: Scikit-learn
        - name: OpenCV, PIL
    - category: Scientific Computing & Data (Expert)
      icon: fa-solid fa-chart-line
      items:
        - name: NumPy, Pandas, Polars
        - name: SciPy, Matplotlib, Seaborn
        - name: Streamlit
    - category: MLOps & DevOps (Master)
      icon: fa-solid fa-infinity
      items:
        - name: Git, GitHub/GitLab CI/CD
        - name: Docker, Singularity
        - name: Kubernetes
        - name: Azure ML, SkyPilot
    - category: Monitoring & Development Tools (Advanced)
      icon: fa-solid fa-chart-bar
      items:
        - name: Weights & Biases
        - name: Pytest, Unittest
        - name: Grafana
    - category: High-Performance Computing (Advanced)
      icon: fa-solid fa-bolt
      items:
        - name: Triton
        - name: SLURM
        - name: Faiss

- title: Research Interests
  type: nested_list
  contents:
    - title: Uncertainty Quantification
      items:
        - Conformal prediction (local, group-conditional, anytime)
        - Distribution-free coverage under concept drift
    - title: Online and Adaptive Learning
      items:
        - Soft sensing for real-time quality monitoring
        - Just-in-Time Learning and similarity-based retrieval
    - title: Explainability and Interpretability
      items:
        - SHAP-based feature attribution
        - Knowledge discovery in industrial processes
    - title: Unsupervised Learning
      items:
        - Self-Organizing Maps
        - Topology-preserving dimensionality reduction

- title: Interests
  type: nested_list
  contents:
    - title: Machine Learning Research
      items:
        - Uncertainty Quantification
        - Diffusion Models
        - Explainability
        - Unsupervised Learning
        - Multimodal AI
    - title: Entrepreneurship & Finance
      items:
        - Investments
        - Product Development
        - Podcasts
        - Creator Economy
    - title: Sports
      items:
        - NBA, MMA, E-sport
        - Weight Training, Swimming
        - Rugby, Handball

- title: Volunteering & Extracurricular
  type: time_table
  contents:
    - title: Handball Player
      institution: University, France & UK
      year: 2020 - 2023
      description:
        - "Achieved 3rd place in the highest national university league (BUCS, UK) and 4th place in the highest national university league (FFSU, France)."
        - "Winner of the 49th Challenge Centrale Lyon and the 19th TRAMS."
    - title: Vice President
      institution: <a href="https://bdsminesales.fr/">Bureau des Sports</a>, Alès, France
      year: 2021 - 2022
      description:
        - "Led a 25-person team to manage 5 tournaments and oversee 20 sports."
        - Secured partnerships with major brands, including <a href="https://www.armorlux.com/">Armor Lux</a> and <a href="https://www.tonyparker.com/">Tony Parker</a>.
        - "Managed the association's website and IT infrastructure."
    - title: Vice President
      institution: <a href="https://www.instagram.com/emavisual/">EmaVisual</a> (Photography & Video), Alès, France
      year: 2021 - 2022
      description:
        - "Led a team of 10 photographers to cover 25 university events."
        - "Edited and produced multiple videos and over 10,000 photos."
    - title: Volunteer
      institution: <a href="https://tsikyzanaka.org/">Tsiky Zanaka</a> (Humanitarian), Alès, France
      year: 2020 - 2021
      description:
        - "Raised over €3500 to fund and support solidarity projects in Madagascar."
        - "Organized and participated in fundraising events, including cooking for 300 students."
    - title: Disc Jockey
      institution: <a href="https://www.instagram.com/emamix_/?hl=fr">EmaMix</a> (DJs Association), Alès, France
      year: 2020 - 2021
      description:
        - "Provided music and technical support for student events and gatherings."
    - title: Logistics Volunteer
      institution: <a href="https://www.instagram.com/trophee_orlandini/">Trophée Orlandini</a> (Rugby Tournament), Alès, France
      year: 2020
      description:
        - "Prepared event logistics and setup for a multi-school rugby tournament."

- title: References
  type: time_table
  contents:
    - title: <a href="https://www.sachahu.com/">Sacha Hu</a>
      maindescription: Senior CV Engineer at Augmodo. Co-supervised my work on uncertainty-aware computer vision for open-world retail deployments. Reach via name at augmodo dot com.
    - title: <a href="https://www.linkedin.com/in/eric-moulines-61512860/">Pr. Eric Moulines</a>
      maindescription: Professor at École Polytechnique and MBZUAI, Member of the French Academy of Sciences. Co-supervised my PhD thesis on conformal and adaptive soft sensors for real-time quality prediction in multistage industrial processes. Reach via name dot surname at polytechnique dot edu.
    - title: <a href="https://www.linkedin.com/in/ahmed-shokry-aa705075/">Dr. Ahmed Shokry</a>
      maindescription: Senior Research Engineer at École Polytechnique. Co-supervised my PhD thesis on conformal and adaptive soft sensors for real-time quality prediction in multistage industrial processes. Reach via name dot surname at polytechnique dot edu.
    - title: <a href="https://www.linkedin.com/in/desroziers/">Dr. Sylvain Desroziers</a>
      maindescription: Manager of the Michelin I4.0 Explore team and Fellow AI at Michelin. Co-supervised my PhD thesis on conformal and adaptive soft sensors for real-time quality prediction in multistage industrial processes. Reach via name dot surname at michelin dot com.
    - title: <a href="https://www.linkedin.com/in/moreaudmaxime/">Dr. Maxime Moreaud</a>
      maindescription: Principal Data Scientist at Michelin. Co-supervised my PhD thesis on conformal and adaptive soft sensors for real-time quality prediction in multistage industrial processes. Reach via name dot surname at michelin dot com.
    - title: <a href="https://www.linkedin.com/in/antoine-cully-3a4a36163/">Pr. Antoine Cully</a>
      maindescription: Professor at Imperial College London and Director of the Adaptive & Intelligent Robotics Lab (AIRL). Supervised my independent research project and my MSc research internship on quality-diversity optimization. Reach via a dot surname at imperial dot ac dot uk.
    - title: <a href="https://www.linkedin.com/in/ludovic-gardy/">Dr. Ludovic Gardy</a>
      maindescription: Senior Data Scientist (Freelance), previously Research Engineer at CNRS. Co-supervised my MSc research internship on ML-based pathological oscillation detection in epilepsy. Reach via name dot g at sotisai dot com.