name: Quentin Velard
role: Machine Learning Engineer
intersection: Unmanned Systems × Machine Learning
education: MSc Applied Math (Mines Nancy) 
currently: Building in stealth · autonomous planning from natural language

Building an autonomous planning system driven by natural language.

Open to conversations on: long-horizon planning, agent architectures, evals for planning, robotics & autonomy interfaces.

ML Engineer focused on building reliable systems end-to-end, with a CV deep in computer vision and industrial R&D (STMicroelectronics, latent diffusion on H100 clusters) and an academic background in quantum machine learning (RMIT, IEEE submission). I turn messy, real-world data into deployable models with clear metrics, robust evaluation, and lightweight MLOps. Now applying that engineering discipline to autonomy and planning.

%%{init: {'theme': 'neutral'}}%%
mindmap
  root((ML Engineering))
    Autonomous Planning
      Natural Language to Plans
      LLM Agents
      Long-horizon Reasoning
      Eval & Robustness
    Computer Vision
      CNN Surrogate Models
      UNET Architectures
      Latent Diffusion
    Industrial ML & R&D
      Reliability Modeling
      POC to Production
      Calibration & Uncertainty
    Quantum ML
      Variational Circuits
      Incremental Learning
      Catastrophic Forgetting
    Generative AI
      GAN
      Fine-tuned LLMs
      Molecule Generation
    MLOps & Distributed Training
      H100 Clusters
      Quantization
      Reproducible Pipelines
    Unmanned Systems & Robotics
      LeRobot Hackathon
      Hardware Tinkering
    VC & Due-Diligence
      Architecture Reviews
      Roadmap Feasibility
      France 2030

Continual learning with variational quantum circuits, studying resilience to catastrophic forgetting in sequential tasks. Co-authored paper under IEEE peer-review.

Deep learning pipeline combining GANs and a fine-tuned LLM to generate novel biomolecules with antioxidant and anti-inflammatory properties. Report available on velard.fr.

Hands-on with the LeRobot stack, exploring the path from hobbyist robotics to autonomous decision-making.

More projects on velard.fr

Member of the Student Scientific Convention on Hydrogen · supporter of the French Nuclear Energy Society (SFEN) · regular attendee of conferences on cryptocurrencies, decentralized finance, AGI, AI for industry, and general models for robotics.

Marathons · swimming · trekking · mountaineering.