John Wilmer Ceron Devia, a systems engineering professional, has released a conceptual design proposal aimed at modernizing IT support ticket management through artificial intelligence and process automation. Published on DEV.to in November 2026, the document addresses the operational inefficiencies caused by scattered request channels such as email, spreadsheets, and informal communications. The core premise is that a centralized, intelligent system can improve classification, traceability, and overall service continuity for organizations struggling with fragmented support workflows.

Architectural Components and Tech Stack

The proposed architecture is built on a modular stack designed for practical implementation. It features a web-based user interface for request registration and consultation, backed by a relational database to store user data, ticket states, categories, and change history. The intelligence layer consists of an AI module that suggests categories based on textual descriptions, while an automation engineβ€”specifically citing tools like n8n or Power Automateβ€”handles notifications and assignments. A Power BI dashboard is recommended for visualizing key metrics such as request volumes and resolution times.

Methodology and Evaluation Metrics

Ceron Devia outlines a five-phase applied research methodology, moving from requirements analysis and architectural design to prototype development, validation, and final evaluation. Since this is a proposal rather than a completed study, no experimental results are presented. Instead, the document defines specific indicators for future assessment, including classification accuracy, macro-F1 scores, average registration time, and the percentage of AI suggestions corrected by human staff. The goal is to compare these metrics against a documented baseline to prove technical viability.

Security and Ethical Considerations

The proposal emphasizes strict security protocols, including role-based access control and authentication mechanisms to protect personal data. A critical ethical guideline suggests avoiding the transmission of unnecessary credentials or sensitive information to external AI models. The design ensures that all AI-generated suggestions remain subject to review and correction by authorized personnel, maintaining human oversight in the decision-making process.

Key Takeaways

  • The system integrates n8n or Power Automate for workflow automation alongside a relational database for structured data storage.
  • Evaluation metrics focus on classification accuracy (macro-F1) and the rate of human corrections to AI suggestions.
  • The project is currently a conceptual proposal with no empirical data, pending future prototype validation.
  • Security guidelines explicitly warn against sending sensitive credentials to external AI APIs.

The Bottom Line

While the architecture sounds solid on paper, the lack of empirical data means this is still just a blueprint. Developers should treat this as a reference for structuring their own AI-augmented helpdesk tools, not as a proven solution.