
Home Institution: Princeton University
Ianna Osborne
Research Software Engineer, Princeton UniversityMy research:
I am a Research Software Engineer specializing in open-source software development within the Python ecosystem and Julia for fundamental physics. My work focuses on designing and optimizing software for high-energy physics (HEP) data analysis.
In the past, I contributed extensively to CMS software (CMSSW), particularly in geometry description, simulation, and event display. My expertise spans software architecture, numerical optimization, and large-scale scientific computing, with a strong focus on enhancing data analysis tools for physics research.
My expertise is:
- Programming Languages: C++, Python, Julia
- Technologies: Awkward Array, Numba, Just-in-Time Compilation
- Integration: Python integration with other languages
- Software Development: Open-source contributions in Python and Julia ecosystems
- Scientific Computing: Geometry modeling, simulation, event visualization
- Configuration & Deployment: Managing and distributing complex software environments
A problem I’m grappling with:
I am currently focused on integrating Awkward Array with Numba to optimize performance through Just-in-Time compilation techniques. Additionally, I am exploring Python on GPU to leverage parallel computing power for complex data analysis.
I’ve got my eyes on:
- High-performance computing (HPC) applications in fundamental physics
- Large-scale parallelism and deep storage hierarchies for HEP data
- Julia’s potential for optimizing scientific computing workflows
I want to know more about:
To further improve my work, I am keen to deepen my understanding of:
- Optimizing data movement and storage for large-scale scientific workflows
- Heterogeneous computing to leverage CPUs, GPUs, and specialized accelerators
- Efficient memory management strategies for structured and unstructured data
- Scalable distributed systems that handle massive physics datasets
By exploring these areas, I aim to contribute to more efficient and scalable computational solutions for HEP.

