Lionel Garcia

en/fr

• Providing scientific software development services to academic and industrial clients. Specializing in data analysis workflows and pipeline development.
• Currently working on astronomy-related projects.

• Led a high-impact scientific program with the James Webb Space Telescope, managing a team of 15 scientists.
• Developed, released, and maintained jaxoplanet (a suite of models to analyze large astronomical time-series datasets), nuance (a Gaussian process algorithm to detect periodic signals in time-series dominated by red noise), and spotter (a framework to stochastically model active stellar surfaces). These tools are powered by JAX, optimized for GPUs/TPUs, and used in the context of Bayesian modeling, inference and machine learning applications.
• Published all research as open-source Python packages with comprehensive documentation, CI/CD, and unit testing.

• Maintained and remotely operated large robotic telescopes for the SPECULOOS project. Developed a fully automated image processing pipeline running daily to transform raw data into scientific products. Built and maintained web dashboards for data analysis monitoring and visualization (40 users).
• Developed, released, and maintained prose (astronomical image processing pipelines), twirl (pattern matching for stellar asterisms), and ballet (CNN for stellar PSF fitting).

• Applied machine learning to collect data and predict costs for future space missions.

• Designed the data analysis workflow for calibrating a Large Hadron Collider instrument.