Administration of a computing cluster for elementary particle physics (June 2019 - Jun 2024)

The ATLAS-BFG cluster is a high-throughput cluster (HTC) that allows users from local particle physics research groups and global members of the ATLAS collaboration to analyse collision data from the ATLAS experiment . The core of the cluster is a batch system with 3400 CPU cores and a distributed object storage system with 4 PB of storage space. The cluster is part of the World Wide LHC Computing Grid (WLCG).

• Participation in the operation and maintenance of the cluster
• Integration of new and existing services into the cluster's IaC solution.
• Maintenance and further development of the cluster's monitoring infrastructure
• Support for local users

Technologies used: Puppet, Packer, Docker, Prometheus, Grafana, Nagios, Slurm, CentOS 7, Alma 9

Framework for the accounting of shared and opportunistic resources (April 2023 - June 2024)

AUDITOR (AccoUnting DatahandlIng Toolbox for Opportunistic Resources) is a flexible and extensible framework for the accounting of shared and opportunistic resources.

• Planing of new features and technical management of contributing developers
• Release management: Publishing new releases, communication with users
• Operation of multiple AUDITOR instances within the ATLAS-BFG cluster

Technologies used: Rust, Python, PostgreSQL, Docker, GitHub Actions

Monitoring of WLCG grid sites (January 2022 - June 2024)

HammerCloud is a monitoring and testing framework, that is developed and used by the ATLAS collaboration to monitor the availability and performance of the global grid infrastructure of the WLCG.

• Migrating from Python 2/Django 1 to Python 3/Django 4
• Migration to CERN's new OIDC-based single sign-on solution
• Redesign of the architecture to a container-based setup

Technologies used: Python, Django, Docker, Docker Compose

PhD Thesis: Search for Higgs‑Boson Pair‑Production in the bbll+ETMiss final state with the ATLAS detector at √s = 13 TeV (February 2018 - May 2023)

• Measurement of the cross section and signal strength of non-resonant Higgs-boson pair-production
• Measurement of the trillinear Higgs-boson self-coupling
• Analysis of ~100TB of data
• Improving the event selection with neural networks
• Statistical analysis (confidence intervals, hypothesis testing)
• Improved sensitivity by a factor of three compared to previous analysis
• Six-month research stay at CERN

Technologies used: Python, Django, Docker, Docker Compose

Teaching duties (April 2018 - June 2024)

• Correction of weekly exercises and final exams
• Discussion of the exercises with the students in weekly tutorials