grafana

To address redundant data scanning and high query latency caused by overlapping windows in time-series monitoring systems, this paper proposes PromSketch—the first “approximation-first” intermediate caching framework. PromSketch innovatively integrates Count-Min Sketch with sliding-window approximate aggregation to construct a pluggable caching layer, augmented by a rule-aware cache eviction policy and a lightweight metadata index. It achieves efficient coverage for 70% of prevalent time-series aggregation queries, with average approximation error ≤5%. Experimental evaluation demonstrates that, compared to Prometheus, PromSketch reduces query latency and computational cost by two orders of magnitude; relative to VictoriaMetrics, it cuts cost by over 4×. This work establishes the first systematic framework enabling low-overhead, high-coverage approximate query processing for windowed time-series monitoring workloads.

This work addresses the challenge developers face in comprehensively evaluating the cost, code quality, and behavioral patterns of AI-assisted programming. The paper presents the first AI observability system grounded in real API invocations, integrating precise token tracking, a multi-provider LLM gateway, a pricing database covering 24 models, a response validation pipeline, and LLM-driven code review analysis. A unified dashboard enables joint insights into cost and quality, while Prometheus-based metrics ensure systematic monitoring. In six months of real-world deployment, the system maintained per-review cost estimation errors within 2% and improved the efficiency of analyzing AI usage patterns by an order of magnitude.

To address insufficient observability in software systems—leading to slow fault localization and low operational efficiency—this paper designs and implements a lightweight, full-stack observability framework. The framework integrates Java bytecode instrumentation with event stream collection to enable runtime call-chain tracing, performance diagnostics, and root-cause analysis. It introduces a novel dual-mode deployment architecture supporting both online services and on-premises deployment, and achieves cross-toolchain collaborative visualization via tight REST API integration with ExplorViz. Evaluated on the TeaStore benchmark, the system delivers millisecond-scale distributed tracing and real-time heatmap rendering, reduces end-to-end latency by 32%, and shortens mean time to fault identification to the minute level. These results significantly enhance observability and operational intelligence for microservice systems.

Cloud-native systems face fragmented observability and challenging root-cause analysis due to their distributed, highly dynamic architectures. To address this, this paper proposes a reusable, observability-oriented design pattern system comprising three core categories: distributed tracing, application-level metric modeling, and infrastructure-level metric collection. Unlike ad-hoc toolchain integrations, our work is the first to systematically abstract industrial practices into structured, composable design patterns that holistically guide end-to-end monitoring architecture. Implemented and validated using mainstream frameworks—including OpenTelemetry and Prometheus—the approach significantly improves latency attribution accuracy, resource utilization assessment efficiency, and anomaly detection timeliness. Empirical evaluation across multiple microservice deployments demonstrates a 42% reduction in mean time to identify failures.

Dashboards rely heavily on text for contextual explanation, insight communication, and interactive guidance; however, existing dashboard authoring tools prioritize visualization while offering limited, ad-hoc support for text creation. To address this gap, we propose the first interactive, human-in-the-loop system specifically designed for dashboard text authoring. Our approach comprises two core components: (1) a taxonomy of dashboard text types coupled with a visual–semantic mapping model that formalizes relationships between visual encodings and textual semantics; and (2) a scaffolded authoring framework integrating rule-based semantic constraints with large language models (LLMs) to ensure generation quality, author controllability, and visual–semantic consistency. The system supports context-aware generation, readability optimization, and layout-aware editing. In an evaluation with 12 professional dashboard authors, our method improved insight communication accuracy by 37% and significantly increased author satisfaction—demonstrating its effectiveness in seamless workflow integration and real-world utility.

This work addresses the inefficiencies in multicenter radiology research caused by reliance on manual communication and shared spreadsheets, which hinder timely data exploration and coordination. The authors propose the first lightweight, open-source monitoring framework tailored for multicenter medical imaging studies, built upon the Grafana-Prometheus stack. By aggregating distributed metrics and offering configurable visualization dashboards, the framework enables privacy-preserving cross-institutional monitoring without sharing raw data. It is deeply integrated into the Kaapana platform and has been deployed across 38 university hospitals within Germany’s RACOON consortium, significantly enhancing transparency and operational efficiency in research coordination. The source code is publicly available.

Existing approaches to monitoring large language models are confined to isolated layers of the system stack and lack a systematic analysis of the interplay and complementarity among full-stack observability techniques. This work proposes the first unified five-layer AI observability framework, integrating cutting-edge advances from MIT, UC Berkeley, OpenAI, and Microsoft Research in areas such as confidence calibration, internal state probing, chain-of-thought monitoring, cloud operations benchmarking, and non-intrusive tracing. Through a structured evaluation of the strengths, limitations, and applicability of techniques across each layer, the study identifies four critical research gaps and highlights the integration of model-level signals with infrastructure anomalies as the central challenge. This framework lays the theoretical foundation for building end-to-end intelligent operations systems for large language models.

This work addresses the limitations of traditional real-time analytics systems, which rely on manually defined queries and struggle to proactively uncover the vast array of potential insights within complex, dynamic data streams. To overcome this, the authors propose a multi-agent architecture that establishes a continuous closed-loop process for autonomous insight discovery, encompassing hypothesis generation, compilation of executable analyses, result validation, and visualization. A key innovation is the introduction of a contract-driven design based on typed intermediate artifacts, which ensures modularity, observability, lineage tracking, and secure execution of dynamic analyses. The system leverages Kafka as its event coordination backbone and Flink for stream processing, integrating large language models to power specialized agents. Empirical evaluations in retail, financial, and public data scenarios demonstrate an effective paradigm shift from query-driven to proactive discovery-driven analytics.

This work addresses the lack of accessible learning and experimentation environments for the GraphAlg language by presenting the first fully browser-based online platform. The platform enables users to write and execute GraphAlg programs directly in the browser without any installation, while integrating interactive tutorials to facilitate onboarding for beginners and rapid prototyping for experts. Built upon modern web frontend technologies, it incorporates a custom GraphAlg interpreter and an in-browser execution engine, unifying educational support and algorithm development within a single interface. The system has been open-sourced and publicly deployed, significantly lowering the barrier to entry for learning GraphAlg and reducing the cost of graph algorithm validation.

Existing approaches struggle to reconstruct interactive data dashboards that support functionalities such as clicking and filtering. This work introduces Dashboard2Code, a novel task requiring models to actively explore interactive dashboards and integrate user interaction feedback to generate code that faithfully reproduces the target dashboard. To facilitate research in this direction, we present DashboardMimic, the first benchmark dataset built on Plotly+Dash, along with an automated evaluation framework that combines semantic analysis and dynamic interaction testing. Experimental results on 180 high-quality dashboard–code pairs demonstrate that current models exhibit limited performance on highly complex dashboards, with closed-source models significantly outperforming their open-source counterparts.