Engineering
new AI
methods.

Comtrade AI is shaped by a research-first culture, maintained through strong, ongoing ties to the academic research community and continuous exposure to current scientific work in the field. This keeps our thinking close to emerging methods and rigorous evaluation practices, and ensures that what we build is grounded in evidence rather than hype.

In particular, Dr. Martin Molan is an active researcher at the University of Bologna, which helps keep Comtrade AI connected to contemporary research discourse and standards. This scientific work is carried out in an academic context as part of independent research activity by our researchers, and is referenced here to demonstrate rigor and capability rather than to imply organizational ownership.

In practical terms, it strengthens our ability to tackle problems where off-the-shelf approaches fail, and to translate high-quality research into defensible, production-grade IP.

• Borghesi, A. et al. Anomaly Detection and Anticipation in High Performance Computing Systems. IEEE Transactions on Parallel and Distributed Systems, 2021.
• Molan, M. et al. RUAD: Unsupervised Anomaly Detection in HPC Systems. Future Generation Computer Systems, 2023.
• Borghesi, A. et al. M100 ExaData: A Data Collection Campaign on the CINECA Marconi100 Tier-0 Supercomputer. Scientific Data, 2023.
• Molan, G. et al. Model for Quantitative Estimation of Functionality Influence on the Final Value of a Software Product. IEEE Access, 2023.
• Molan, M. et al. GRAAFE: Graph Anomaly Anticipation Framework for Exascale HPC Systems. Future Generation Computer Systems, 2024.
• Jati, G. et al. LIDAROC: Realistic LiDAR Cover Contamination Dataset for Enhancing Autonomous Vehicle Perception Reliability. IEEE Sensors Letters, 2024.
• Khan, J. A., Molan, M., Bartolini, A. EXASAGE: The First Data Center Operational Data Analysis Assistant. Future Generation Computer Systems, 2026.

• Molan, M. et al. An Explainable Model for Fault Detection in HPC Systems. International Conference on High Performance Computing, 2021.
• Molan, M. et al. Semi-Supervised Anomaly Detection on a Tier-0 HPC System. ACM International Conference on Computing Frontiers, 2022.
• Molan, M. et al. Analysing Supercomputer Nodes Behaviour with the Latent Representation of Deep Learning Models. European Conference on Parallel Processing, 2022.
• Molan, M. et al. Machine Learning Methodologies to Support HPC Systems Operations: Anomaly Detection. European Conference on Parallel Processing, 2022.
• Molan, M. et al. Graph Neural Networks for Anomaly Anticipation in HPC Systems. ACM/SPEC International Conference on Performance Engineering, 2023.
• Khan, J. A. et al. The Graph-Massivizer Approach Toward a European Sustainable Data Center Digital Twin. IEEE COMPSAC, 2023.
• Molan, M. et al. ExaQuery: Proving Data Structure to Unstructured Telemetry Data in Large-Scale HPC. ACM/SPEC International Conference on Performance Engineering, 2024.
• Guindani, B. et al. Exploring the Utility of Graph Methods in HPC Thermal Modeling. ACM/SPEC International Conference on Performance Engineering, 2024.
• Jati, G., Molan, M. et al. AutoGrAN: Autonomous Vehicle LiDAR Contaminant Detection Using Graph Attention Networks. ACM/SPEC ICPE Companion, 2024.
• Jati, G., Molan, M. et al. TinyLid: A RISC-V Accelerated Neural Network for LiDAR Contaminant Classification in Autonomous Vehicles. ACM International Conference on Computing Frontiers, 2024.
• Jati, G., Molan, M. et al. AutoGrAN Extended Abstract. Robotics: Science and Systems Workshop on Semantics for Robotics, TU Delft, 2024.
• Jati, G., Molan, M. et al. LIDAROC: Realistic LiDAR Cover Contamination Dataset for Enhancing Autonomous Vehicle Perception Reliability. IEEE Sensors Conference, 2024.
• Jati, G., Molan, M. et al. False Confidence: Is Detection Reliable Enough for Autonomous Vehicles in LiDAR Corruption? IEEE Sensors Conference Workshop, 2024.
• Jati, G. et al. Ensemble Contaminant Classification for LiDAR, 2025.
• Khan, J. A. et al. ExaPrompt: The First ODA Co-Pilot, 2025.

• Dagstuhl Seminar 25171: Holistic Graph-Processing Systems: Enabling Real-World Scale and Societal Impact, 2025.
• Availability Is Sustainability, DATE 2024 Special Day.
• How AI Ensures Optimal Operation of Data Centres and Supercomputers, Expo 2020 Dubai, Slovenian Pavilion.
• Unlocking Unusual Data: Particles, Biology, Space, AI_4_LIFE BioTech Future Forum, Belgrade 2024.

• AI-Powered Resilient Perception for Autonomous Systems, PhD Co-Supervision, 2025.
• Graph Neural Networks for Anomaly Anticipation in HPC Systems, MSc Co-Supervision, University of Bologna, 2023.
• Deep Learning Approach to Modelling and Forecasting a Tier-0 Data Center's Power Usage Effectiveness, MSc Co-Supervision.