Industrial data scientists require deep domain understanding to model asset conditions effectively, yet traditional sources such as Subject Matter Experts (SMEs) and Failure Modes and Effects Analysis (FMEA) documents are often unavailable or incomplete. We present a deployed Multi-Agent System (MAS) that leverages Large Language Models (LLMs) to automatically generate and refine domain-relevant questions, improving modeling decisions across industrial projects. The system addresses two key challenges—ensuring linguistic diversity and maintaining high relevance—by combining established information diversity metrics with a grounded relevancy classifier. We evaluate its effectiveness through diversity benchmarks, compare against direct prompting and AutoAgents baselines, knowledge coverage on downstream FMEA tasks, and controlled user studies. Deployed in real-world projects, the MAS has improved multiple stages of the CRISP-DM methodology, resulting in measurable savings in cost and man-hours.

We introduce FailureSensorIQ, a novel Multi-Choice Question-Answering (MCQA) benchmarking system designed to assess the ability of Large Language Models (LLMs) to reason and understand complex, domain-specific scenarios in Industry 4.0. Unlike traditional QA benchmarks, our system focuses on multiple aspects of reasoning through failure modes, sensor data, and the relationships between them across various industrial assets. Through this work, we envision a paradigm shift where modeling decisions are not only data-driven using statistical tools like correlation analysis and significance tests, but also domain-driven by specialized LLMs which can reason about the key contributors and useful patterns that can be captured with feature engineering.We evaluate the Industrial knowledge of over a dozen LLMs including GPT-4, Llama, and Mistral on FailureSensorIQ from different lens using Perturbation-Uncertainty-Complexity analysis, Expert Evaluation study, Asset-Specific Knowledge Gap analysis, ReAct agent using external knowledge-bases.Even though closed-source models with strong reasoning capabilities approach expert-level performance, the comprehensive benchmark reveals a significant drop in performance that is fragile to perturbations, distractions, and inherent knowledge gaps in the models.We also provide a real-world case study of how LLMs can drive the modeling decisions on 3 different failure prediction datasets related to various assets.We release: (a) expert-curated MCQA for various industrial assets, (b) FailureSensorIQ benchmark and Hugging Face leaderboard based on MCQA built from non-textual data found in ISO documents, and (c) ``LLMFeatureSelector'', an LLM-based feature selection scikit-learn pipeline. The software is available at https://github.com/IBM/FailureSensorIQ.