Reasoning-Aware LLMs for Transparent, Explainable Workflow Automation

Mondorf & Plank (2024) point out that most LLM research focuses on output accuracy, with little emphasis on explicability or reasoning transparency—especially in workflow contexts. This idea proposes developing LLM-powered workflow automation systems that are “reasoning-aware”: for each automated decision (such as data transformation, schema matching, or incident triage), the system provides a natural language rationale, cites relevant data or policies, and allows for user queries or challenges. This not only builds trust and accountability—crucial in regulated domains like healthcare (Berry et al., 2025)—but also facilitates debugging and continuous improvement. By systematically comparing reasoning-aware LLMs to standard “black box” workflow automation, this research could set new standards for transparency and auditability in automated data management.

References:

1. Utilizing large language models for gastroenterology research: a conceptual framework. Parul Berry, R. Dhanakshirur, S. Khanna (2025). Therapeutic Advances in Gastroenterology.
2. Beyond Accuracy: Evaluating the Reasoning Behavior of Large Language Models - A Survey. Philipp Mondorf, Barbara Plank (2024). arXiv.org.