Observability Data Centralization and Tracing Implementation

• Rishab Kumar Jha

Room: 503/1-001 - Council Chamber **Project Description**: Built a proof-of-concept for centralized observability and distributed tracing across multiple Kubernetes clusters. Unified metrics, logs, and traces into a single Grafana dashboard where you can pick any cluster and service and correlate all three, including jumping from a latency spike straight to the exact trace. Designed a scalable parent/child OpenTelemetry architecture in which each child cluster forwards spans to a central gateway that routes every trace consistently and tail-samples them, then stores them in Jaeger and OpenSearch. Evaluated instrumentation approaches (auto-instrumentation vs SDK), and took the pipeline toward production: measuring resource cost, reliability, and security, adding self-monitoring for the pipeline itself, and extending tracing to the edge with a native OpenTelemetry filter in HAProxy that can turn tracing on and off per service at runtime. **Supervisor**: Daniel Gonçalves Portovedo, Antonio Nappi

Oracle SQLcl MCP server evaluation and resource limits

• David Benko

Room: 503/1-001 - Council Chamber **Project Description**: Evaluate and benchmark Oracle SQLcl Model Contect Protocol (MCP) server performance and accuracy using open agentic models. Document installation and use of Oracle SQLcl MCP server for popular developer tools (e.g. VSCode). Implement automatically applied DB resource limits for database sessions identified as running an AI/MCP based client. **Supervisor**: Andrzej Nowicki, Miroslav Potocky

Implementation of a modern logging pipeline for Hadoop

• Jordie Biemold

Room: 503/1-001 - Council Chamber **Project Description**: The Hadoop infrastructure at CERN consists of many services (like HDFS, YARN, Spark, ZooKeeper) that all collect logging data in their own files. These services are used to store large amounts of data (from experiments and other sources) and process it efficiently. Before, the logs were located on each server individually, this made looking for (the source of) errors difficult. The goal of the project was to centralise the collection of logging data across all services using Fluent Bit (deployed through Puppet and routing through CERN MONIT infrastructure). We use OpenSearch as a storage backend and visualise the data in Grafana dashboards. It will help Hadoop system administrators investigate incidents faster and detect problematic service behaviour before it becomes a problem. **Supervisor**: Luis Pigueiras

Oracle performance and availability alerts in Grafana

• Kilian Mahler

Room: 503/1-001 - Council Chamber **Project Description**: The goal of this project is to evaluate and implement monitoring and alerting for Oracle databases using CERN’s centralized monitoring infrastructure, [MONIT](https://monit.docs.cern.ch/), together with the Oracle Database Metrics Exporter. **Supervisor**: Abel Cabezas Alonso, Miroslav Potocky

Numerical study of floating point precision within the PEPPER event generator

• Yucheng Zhang

Room: 503/1-001 - Council Chamber **Project Description**: This project investigates whether PEPPER can safely use reduced floating-point precision to accelerate parton-level event generation on modern CPUs and GPUs. By comparing FP32 and FP64 calculations with CADNA, it identifies where rounding errors affect matrix elements and event weights, supporting the development of a reliable mixed-precision approach for HL-LHC simulations. **Supervisor**: Dr. Enrico Bothmann & František Stloukal

GPU implementation of tensor network states

• Aditya Tandon

Room: 503/1-001 - Council Chamber **Project Description**: The main aim of this project was to identify suitable platforms for doing Tensor Network simulations of entangled quantum systems and to leverage massively parallel GPUs to speed them up. For this purpose, benchmarking against existing thesis codes and libraries must be performed before testing the simulations on research systems. **Supervisor**: Dr Enrique Rico Ortega

Performance Study of GPUDirect Storage for CMS Reconstruction Workflows

• Ayaan Aziz Dewani

Room: 503/1-001 - Council Chamber **Project Description**: Modern high-energy physics experiments such as CMS increasingly rely on heterogeneous computing architectures combining CPUs, GPUs, and high-performance storage systems. While GPU acceleration has significantly improved compute throughput, I/O and data movement between storage, host memory, and device memory remain a critical bottleneck. GPUDirect Storage (GDS) is a recent technology that allows GPUs to directly access data from NVMe storage, bypassing the CPU and reducing latency and memory copies. The goal of this project is to study, benchmark, and optimize the use of GPUDirect Storage and to evaluate how GDS impacts throughput and latency compared to traditional I/O paths using host memory. **Supervisor**: Markus Holzer, Felice Pantaleo

Fast Decision Forest inference on AMD AI Engines

• Ayoub Agouzoul

Room: 503/1-001 - Council Chamber **Project Description**: **Supervisor**: Sioni Paris Summers, Ioannis Xiotidis

Mojo: the future programming language for CERN experiments?

• Abbas Naim

Room: 503/1-001 - Council Chamber **Project Description**: **Supervisor**: Axel Naumann

Trigger Level Analysis data preparation through co-processors

• Emilly Rodrigues (CERN)

Room: 503/1-001 - Council Chamber **Project Description**: The project studies a compact machine-learning model for reconstructing jets directly from six-layer ATLAS calorimeter images. The target environment is the FPGA-based Global Event Processor of the Phase-II Global Trigger, where the reconstruction must preserve the physically relevant jet information while using a small model. The work therefore evaluates reconstruction quality together with model size and hardware-oriented constraints. **Supervisors**: Ioannis Xiotidis and Noah Clarke Hall

Accelerating ATLAS Muon Tracking with Machine Learning-Based Segment Classification

• Karla Lucic (CERN)

Room: 503/1-001 - Council Chamber **Project Description**: An attention-based graph neural network that filters muon-segment pairs before track seeding in the ATLAS Phase-II reconstruction, validated end to end in ATHENA. **Supervisor**: Davide Di Croce, Johannes Junggeburth

MLOps and Continual Learning for LHC Hardware Triggers

• Maksymilian Artur Moskwytyn (CERN)

Room: 503/1-001 - Council Chamber **Project Description**: This project develops a reproducible, hardware-aware MLOps pipeline for optimizing neural-network models for LHC hardware triggers. It combines automated architecture and quantization search, experiment tracking, physics-level validation, and FPGA deployment, with a reusable workflow demonstrated on an MHA-based minimum-bias anomaly-detection use case. **Supervisor**: Ioannis Xiotidis

GPU-Based Muon Segment Fitting for HL-LHC Reconstruction.

• Manfred Przemyslaw Gawlas (CERN)

Room: 503/1-001 - Council Chamber **Project Description**: Muon reconstruction is a central component of the ATLAS physics program and a key contributor to the overall computational cost of event reconstruction. The muon spectrometer provides precision tracking in the outermost region of the detector, and muon segment fitting is a core step in the outside-in reconstruction chain. Our group has implemented a highly optimized CUDA-based $\chi^2$ minimization fitter. However, approximately half of the current runtime is spent on host–device memory transfers, limiting suitability for latency-critical applications such as triggering. The proposed project aims to build on this foundation by developing a deeper GPU-resident reconstruction pipeline in which additional steps of the muon reconstruction chain are executed directly on the GPU. By moving hit preparation, region construction, and segment seeding onto the device, the project seeks to minimize data movement and amortize memory transfer costs. The reconstruction will be integrated more deeply with the A Common Tracking Software (ACTS) framework using a realistic ATLAS detector geometry and more complex physics processes, including jets and pileup. A full benchmarking campaign will be carried out to compare the GPU-based pipeline with the existing CPU implementation and to assess the feasibility of GPU acceleration for later-stage triggering and quasi-online reconstruction. **Supervisor**: Johannes Junggeburth, CERN

Optimisation of the ATLAS HL-LHC trigger menu for improved diHiggs (bbtautau) measurement

• Sergio Estrada Paredes (CERN)

Room: 503/1-001 - Council Chamber **Project Description:** This project focuses on optimising the ATLAS HL-LHC trigger menu for Higgs-pair production, targeting the HH -> bbτ(had)τ(lep) final state. The goal is to maximise the signal efficiency while maintaining the background rate within the available trigger budget. To achieve this, the muon and electron channels are optimised separately across the trigger chain. At Level-1, different object definitions and preselections are studied, combining large grid scans with BDTs to identify efficient and interpretable selections. The analysis is then extended to the Event Filter, where b-tagging and additional identification requirements provide further background rejection. **Supervisors:** Christopher Young and Tamara Vázquez Schröder

Extending the hls4ml Backend for PandA/Bambu

• Shahd Abuhassan

Room: 503/1-001 - Council Chamber **Project Description**: Hls4ml is an open-source toolkit that converts trained machine-learning models into FPGA firmware using high-level synthesis, enabling ultra-low-latency inference for real-time applications such as LHC trigger systems. PandA/Bambu is an advanced open-source HLS framework that compiles C/C++ into optimized hardware designs and offers a rich set of optimization passes and backend targets. In this project, the student will develop a new hls4ml backend that generates PandA/Bambu-compatible C++ and integrates Bambu’s HLS flow into the hls4ml build pipeline. The work includes implementing dataflow-friendly code generation, tuning Bambu optimizations for ML workloads, and benchmarking performance and resource usage on FPGA targets. This project provides hands-on experience with HLS, ML inference acceleration, and open-source hardware design tools. **Supervisor**: Nicolò Ghielmetti, George Flengas, Dimitrios Danopoulos

Optimized C++ Inference in SOFIE for Quantized Machine Learning Models

• Shaun Christopher Lee

Room: 503/1-001 - Council Chamber **Project Description**: Machine learning models are widely used at CERN for a variety of tasks, including data analysis, pattern recognition, and anomaly detection. These models are often highly sophisticated in order to handle the complexity of the data they process. In addition, when deployed in real-time systems such as trigger pipelines, they must deliver extremely fast inference performance. With the upcoming upgrades of the LHC experiments for the High-Luminosity phase, machine learning models are expected to become even more complex while simultaneously needing to process significantly higher data rates. This places stringent requirements on inference latency, memory usage, and hardware efficiency. SOFIE (System for Optimized Fast Inference code Emit) is a tool that translates trained machine learning models into optimized C++ code with BLAS as its only dependency. In its current implementation, SOFIE-generated code has demonstrated up to 2× faster inference and approximately 50% lower memory consumption on CPUs compared to ONNX Runtime for commonly used models in high-energy physics research. SOFIE is also being extended to support inference on heterogeneous architectures through alpaka, an abstraction library for portable heterogeneous programming. In its next iteration, SOFIE will have integrations with PQuant, a library for end-to-end hardware-aware model compression using pruning and quantization techniques. This integration will enable the generation of C++ inference code from quantized machine learning models. **Supervisor**: Sanjiban Sengupta, Lorenzo Moneta

ML model optimization pipeline development

• Stylianos Kostakis (Aristotle University of Thessaloniki (GR))

Room: 503/1-001 - Council Chamber **Project Description**: Machine learning is rapidly replacing heuristic-based algorithms across many domains, and ATLAS plans to make extensive use of ML in both its hardware and software triggers during the HL-LHC era. While mature optimization platforms exist for deploying ML in software, very few MLOps pipelines targeting custom FPGA-based electronics have been tested at scale. Within Next-Generation Triggers (NGT) WP2.1, we are developing a dedicated MLOps pipeline that automatically optimizes and deploys ML algorithms for the ATLAS L0-Global hardware trigger. A prototype version already exists and is in use, while a more generalized approach, one that can potentially extend beyond the specific scope of WP2.1, is being developed in parallel. This work covers three areas: generating open data samples for validating pipeline developments; developing a command-line interface that allows users to create standalone ML algorithms, integrate them into the pipeline, and deploy custom pipeline stages; and integrating a software stack dedicated to the joint optimization of multiple trigger algorithms. The goal is a service-level design through which ML developers for hardware triggers can easily obtain deployable, optimized trigger algorithms that efficiently utilize the underlying hardware. Such a service significantly lowers the barrier to integrating ML into the ATLAS hardware trigger and ensures that common, well-defined metrics are used to evaluate whether deploying an ML model there provides a real advantage. **Supervisor**: Ioannis Xiotidis, Noah Clarke Hall

NGT: Physics Performance Evaluation of Continuous Real-Time Calibrations within the CMS High-Level Trigger System

• Maximilian Spannring

Room: 503/1-001 - Council Chamber **Project Description**: Accurate physics reconstruction of collision events at the LHC relies on precise knowledge of detector conditions at the moment an event is acquired. As these conditions fluctuate during data-taking, calibrations must be derived to account for them. Within the Next Generation Triggers (NGT) project, a novel approach is being tested to derive calibrations for online reconstruction at the High-Level Trigger (HLT) with much higher frequency than current operations. A demonstrator has been commissioned to test the feasibility of this new paradigm for the Phase-2 conditions framework, with the aim of deriving calibrations continuously within a strict 8-hour window while using limited computing resources and maintaining physics quality. This summer student project will focus on a retrospective, comprehensive physics validation of this demonstrator using results collected during the final Run 3 data-taking period. The student will perform a comparative analysis between the conditions derived by the NGT Demonstrator, the standard HLT conditions, and the high-precision Offline conditions which serve as the reference truth. The goal is to quantify improvements in physics object reconstruction and assess triggering efficiencies, to determine if the continuous calibration approach is feasible for Phase-2. **Supervisor**: Jessica Prendi, Marco Musich

Evaluating Jet-Plus-Tracks Reconstruction for the CMS NGT Scouting Stream

• Weronika Karolina Sobien

Room: 503/1-001 - Council Chamber **Project Description**: The CMS experiment relies on rapid event reconstruction at the High-Level Trigger to select interesting proton–proton collisions in real time. For the upcoming Run 4, the Next Generation Trigger (NGT) scouting stream aims to record events at exceptionally high rates by utilizing dedicated fast reconstruction algorithms and reduced data formats. This presentation evaluates the feasibility of the Jet-Plus-Tracks (JPT) algorithm as a lightweight alternative to the computationally intensive Particle Flow (PF) reconstruction. Using simulated Phase-2 CMS data, the study compares both approaches in terms of basic physics quantities, such as jet energies and event activity, as well as their computing footprint, specifically execution runtime and data size. The results highlight the trade-offs between physics performance and computing costs in the context of the Phase-2 scouting ecosystem. **Supervisor**: Marco Musich, Marco Rovere

Accelerate the CMS electron seeding with extended pixel tracks

• Hannah Grace Havel

Room: 503/1-001 - Council Chamber **Project Description**: With the start of the High Luminosity LHC era, or “phase-2,” the upgraded machine will provide unprecedented volumes of data at extreme energies and intensities. Extracting meaningful physics requires advanced data-analysis techniques and algorithms. Within the CMS experiment, we are developing an innovative online trigger system to handle the enormous data rates expected in phase-2. This will be complemented by a novel Next-Generation Trigger (NGT) Scouting strategy, which records physics events at the full input rate while storing a reduced set of reconstructed physics objects. This approach aims to provide access to rare and unconventional signatures that would otherwise be inaccessible due to data-rate and storage limitations. In the CMS Tracker, we are exploring the use of phase-2 pixel tracks as seeds for the Gaussian Sum Filter (GSF) tracking algorithm. Currently, trajectory seeds are computed on-the-fly using tracker and calorimeter information, despite existing tracks being available. Historically, this was due to the low efficiency of pixel tracks as GSF seeds. Our goal is to assess the feasibility of using phase-2 pixel tracks, particularly NGT’s recently developed extended pixel tracks, as GSF seeds under different track-quality criteria. If GSF physics performance is maintained, eliminating trajectory-seed computation could significantly accelerate the CMS HLT, making this a high-impact contribution. **Supervisor**: Bruno Alves, Wahid Redjeb

NGT WP3 Representation Learning for LHC Triggers

• Estefania Moreno Leon

Room: 503/1-001 - Council Chamber **Project Description**: The objective is to use Machine Learning to derive optimal representations of LHC collision events that naturally cluster Standard Model processes while maximizing sensitivity to anomalous signatures beyond the Standard Model (BSM). The project will focus on applying state-of-the-art representation learning methods, including unsupervised (Autoencoder) and self-supervised (BERT, JEPA, VICReg, DINO) learning, to compress high-dimensional LHC event data into descriptive latent spaces for trigger decisions. You will evaluate existing anomaly-detection algorithms and their latent-space representations, design custom training strategies to produce optimal representations, and analyze them in terms of their underlying statistics and correlations with physical observables. Finally, you will leverage NGT resources to study scaling laws and understand how representations evolve with training sample size and model capacity. **Supervisor**: Sabrina Giorgetti, Maciej Mikolaj Glowacki

Bi-Directional Isaac Sim and MuJoCo Bridge for 3D Digital Twin Powered LHC Robotic Interventions

• Ogulcan Gurelli (OpenLab)

Room: 503/1-001 - Council Chamber This project builds a live, bi-directional bridge between two robotics simulators, MuJoCo and NVIDIA Isaac Sim, to support robotic maintenance interventions inside a 3D digital twin of the LHC. MuJoCo acts as the single source of truth for physics, while Isaac Sim mirrors the simulation in real time, rendering it photorealistically inside CERN's actual accelerator geometry; control commands flow back the other way, so a robot can be driven through the tunnel live. The motivation is practical, maintenance in the LHC tunnels is slow, costly, and constrained by radiation limits on human access, so being able to test whether a robot can reach and operate on a target, without collisions before it ever enters the real tunnel is valuable. Alongside the bridge, the project implements collision detection against the tunnel's CAD geometry and demonstrates general applicability by driving a quadruped robot through the twin with a game controller. Furthermore there is VR-based control, allowing an operator to step inside the twin and rehearse an intervention before anyone, or any robot, enters the tunnel.: **Supervisor**: Drew Mason-Laurence

Isaac Sim Controls for 3D Digital Twin Powered Robotics in the LHC and FCC

• Jeremi Machniewski

Room: 503/1-001 - Council Chamber **Project Description**: **Supervisor**: Drew Mason-Laurence

Certificate handout ceremony Room: Glass-box Restaurant 1