Applied Research
in AIOps.

TheMonitoring.AI is an applied research platform exploring predictive monitoring, operational intelligence, and AI-assisted decision-making.

Research Overview

TheMonitoring.AI was conceived as a research-backed operational platform bridging academic research and real-world IT operations. The platform enables continuous experimentation with machine learning models, time-series forecasting, and operational behavior analysis under realistic infrastructure constraints.

Methodologically, the research combines time-series forecasting, supervised learning, and qualitative analysis of operational feedback within a controlled experimental environment.

Current Research Status

Preliminary Results: Initial evaluations indicate the inherent complexity of real-world operational forecasting. Early Prophet-based models achieved approximately 28.3% accuracy over 2,701 evaluated predictions, providing valuable insights into real-world operational forecasting and identifying opportunities for model improvement through feature engineering, extended training periods, and richer contextual data.

Research Questions

  1. Prediction Accuracy: Can time-series models (Prophet, LSTM) predict incident recurrence with measurable accuracy using at least six months of historical operational data?
  2. Prediction Horizon: What prediction window (1h, 6h, 24h) offers the best balance between accuracy and operational actionability?
  3. Feature Contribution: Which factors (temporal patterns, host behavior, incident frequency) contribute most to predictive performance?
  4. Explainability: Can AI-assisted explanations improve operator trust and understanding of predicted incidents?
  5. Human-in-the-Loop: How does structured operator feedback influence model performance over time in production-like environments?

Datasets

The research is based on a proprietary operational telemetry dataset collected from a real monitoring environment.

Reproducibility: Methodology, system architecture, and experimental protocols will be documented in associated publications to enable independent validation and comparative studies.

Anonymized dataset samples may be made available for academic collaboration and comparative studies upon request.

Research & Publications

TheMonitoring.AI is supported by applied research, industry publications, and thought leadership contributions exploring the future of AIOps, predictive monitoring, and operational decision intelligence.

Research Preprint

Ontology-Driven Explainable AIOps: A Conceptual Framework for Semantic Governance of Autonomous IT Operations

Lucas Renan Vicente Bandeira ORCID

Zenodo Preprint · March 2026

DOI: 10.5281/zenodo.19025889

This paper proposes a conceptual ontology-driven governance layer for AIOps systems, addressing the Semantic Governance Gap — the structural absence of machine-processable semantic traceability in current AIOps architectures.

View Preprint (Zenodo DOI) Download PDF

Citation:
Vicente Bandeira, L. R. (2026). Ontology-Driven Explainable AIOps: A Conceptual Framework for Semantic Governance of Autonomous IT Operations. Zenodo. https://doi.org/10.5281/zenodo.19025889

Research Preprint

Predicting Incident Recurrence in Real-World IT Operations

Lucas Renan Vicente Bandeira ORCID

Zenodo Preprint · December 2025

DOI: 10.5281/zenodo.17954627

This paper presents an empirical study on incident recurrence prediction using real operational monitoring data, highlighting the challenges, limitations, and practical implications of AIOps in production-like environments.

View Preprint (Zenodo DOI) Download PDF

Citation:
Vicente Bandeira, L. R. (2025). Predicting Incident Recurrence in Real-World IT Operations. Zenodo. https://doi.org/10.5281/zenodo.17954627

Industry Article

AIOps: inteligência aplicada à monitorização de infraestruturas críticas

AIOps: intelligence applied to the monitoring of critical infrastructures

Lucas Renan Vicente Bandeira

RTI – Revista de Tecnologia da Informação · January 2026

This article explores how AIOps extends traditional infrastructure monitoring by correlating metrics, logs, and operational events to reduce alert noise, accelerate incident diagnosis, and support more informed operational decisions. Rather than promoting unrestricted automation, the article presents AIOps as an intelligence layer that helps operators manage complex data center environments with greater resilience and efficiency.

Read Article

Book

AIOps Beyond Governance

Lucas Renan Vicente Bandeira

Amazon KDP · 2026

This book explores the evolution of AIOps beyond traditional automation, arguing for a broader framework centered on governance, operational intelligence, and decision support in complex IT environments.

View on Amazon

Academic Collaboration

TheMonitoring.AI is an operational intelligence platform designed to help organizations transform monitoring data into risk-aware operational decisions.

The platform also supports ongoing academic research in AIOps, predictive monitoring, and human-AI collaboration in IT operations.

Academic collaboration inquiries: research@themonitoring.ai