Przemek Pospieszny

Przemek Pospieszny, Dominika P. Brodowicz

In recent years, advances in artificial intelligence (AI), particularly generative AI (GenAI) and large language models (LLMs), have made human-computer interactions more frequent, efficient, and accessible across sectors ranging from banking to healthcare. AI tools embedded in digital devices support decision-making and operational management at both individual and organizational levels, including resource allocation, workflow automation, and real-time data analysis. However, the prevailing cloud-centric deployment of AI carries a substantial environmental footprint due to high computational demands. In this context, this paper introduces the concept of agentic environments, a sustainability-oriented AI framework that extends beyond reactive systems by leveraging GenAI, multi-agent systems, and edge computing to reduce the environmental impact of technology. Agentic environments enable more efficient resource use, improved quality of life, and sustainability-by-design, while simultaneously enhancing data privacy through decentralized, edge-driven solutions. Drawing on secondary research as well as primary data from focus groups and semi-structured interviews with AI professionals from leading technology companies, the paper proposes a conceptual framework for agentic environments examined through three lenses: the personal sphere, professional and commercial use, and urban operations. The findings highlight the potential of agentic environments to foster sustainable ecosystems through optimized resource utilization and strengthened data privacy. The study concludes with recommendations for edge-driven deployment models to reduce reliance on energy-intensive cloud infrastructures.

Przemek Pospieszny, Wojciech Mormul, Karolina Szyndler et al.

Modern software systems generate extensive heterogeneous log data with dynamic formats, fragmented event sequences, and varying temporal patterns, making anomaly detection both crucial and challenging. To address these complexities, we propose ADALog, an adaptive, unsupervised anomaly detection framework designed for practical applicability across diverse real-world environments. Unlike traditional methods reliant on log parsing, strict sequence dependencies, or labeled data, ADALog operates on individual unstructured logs, extracts intra-log contextual relationships, and performs adaptive thresholding on normal data. The proposed approach utilizes a transformer-based, pretrained bidirectional encoder with a masked language modeling task, fine-tuned on normal logs to capture domain-specific syntactic and semantic patterns essential for accurate anomaly detection. Anomalies are identified via token-level reconstruction probabilities, aggregated into log-level scores, with adaptive percentile-based thresholding calibrated only on normal data. This allows the model to dynamically adapt to evolving system behaviors while avoiding rigid, heuristic-based thresholds common in traditional systems. We evaluate ADALog on benchmark datasets BGL, Thunderbird, and Spirit, showing strong generalization and competitive performance compared to state-of-the-art supervised and unsupervised methods. Additional ablation studies examine the effects of masking, fine-tuning, and token positioning on model behavior and interpretability.