I design applied AI systems, exploratory computational projects, and learning technologies that make complex information easier to understand and act on.
My work sits at the intersection of AI engineering, data science, technical education, and research-oriented prototyping. I care about systems that are clear, inspectable, human-centered, and useful in real operational contexts.
A portfolio designed around three compatible directions: practical AI systems, computational research, and technical education.
LLM workflows, decision-support tools, automation pipelines, triage systems, and human-in-the-loop applications for real tasks.
Exploratory notebooks, scientific data pipelines, anomaly detection, visualization, and research-style reports for complex domains.
Project-based learning systems, AI tutors, curriculum design, mentoring, and explanations that help learners move from theory to implementation.
Representative work across applied generative AI, business workflows, decision support, human-in-the-loop systems, technical education, scientific exploration, and data visualization.
A French Streamlit demonstrator for business teams showing how generative AI can transform scattered monitoring signals into a structured strategic briefing, with LLM analysis, source traceability, human validation, and lightweight AI governance.
View repository →An AI-assisted short-film pipeline built with Python and Streamlit. The app turns a magical story premise into a structured production workflow: story package, storyboard, character reference cards, generated images, visual animatic, optional Sora clips, and final assembled video.
View repository → Watch generated video → Read project write-up → View Zenodo archive →A governed AI decision-support architecture for anomaly triage in safety-critical aerospace contexts, combining system integration, escalation logic, and human oversight.
View repository →An AI lead qualification and automation pipeline with structured intake, scoring, LLM-assisted analysis, and safer workflow design.
View repository →An AI-assisted recruitment decision-support tool that compares CVs with job descriptions, extracts evidence, generates recruiter briefings, and supports structured candidate evaluation through a transparent human-in-the-loop workflow.
View repository →A multi-agent client operations router for classifying inbound requests, prioritizing actions, and supporting safer operational triage.
View repository →A curriculum-based tutoring system for French civic exam preparation, combining deterministic scoring, AI explanations, and learner feedback.
View repository →A Tableau data visualization project exploring planetary systems beyond the Solar System through accessible, visually coherent analytics.
Explore on GitHub →An exploratory computational research project using latent spaces, scientific representation, and disciplined speculation to investigate unresolved structure.
Explore on GitHub →Independent, project-based research focused on AI systems for exploration, uncertainty, and scientific attention.
I develop notebook-based research artifacts, technical reports, and reproducible project repositories around AI-assisted discovery, exploratory data systems, and governed decision architectures.
This research orientation supports, rather than replaces, practical work: the goal is to build systems that can be inspected, explained, reused, and taught.
Technical education is not separate from my AI work. It is one of the ways I test whether systems and ideas are actually understandable.
Open for future opportunities.
Modular workflows, orchestration, routing, evaluation, and human oversight.
Exploratory systems for astronomy, bioacoustics, scientific datasets, and frontier detection.
AI tutors, adaptive feedback, technical pedagogy, and mission-based learning environments.