Energy utilities still rely on long-lived enterprise asset management platforms for engineering work management, procurement, and inventory processes. Replacing these platforms is often cost-prohibitive and operationally disruptive, necessitating practical improvement layers. This paper introduces a retrieval assistant designed to enhance day-to-day knowledge access across three operational modes: vendor documentation Q&A, operational data store (ODS) schema Q&A, and user interface usage Q&A.
The runtime method integrates intent understanding, query rewriting, hybrid semantic and vector retrieval, context engineering under token limits, grounded answer generation, and deterministic hyperlink conversion for panel identifiers and cited documents. The data preparation pipeline emphasizes semantic enrichment as the primary quality lever by adding table and field descriptions, normalizing acronyms across sources, and indexing useful row-level context.
A measured pilot shows consistent improvements in retrieval quality and user outcomes. Precision at five improved from 0.56 to 0.72, mean reciprocal rank from 0.43 to 0.58, and normalized discounted cumulative gain (nDCG) at five from 0.51 to 0.66. Median task completion time dropped from 14.2 minutes to 8.3 minutes, while usefulness and confidence both increased to 4.0 on a five-point scale. Although results are based on a small sample and reported as pilot findings, they indicate that intent understanding and semantic enrichment can deliver significant operational value in legacy environments while establishing reusable foundations for future analytics and automation tools.
Blogger's Review: This paper showcases how an intelligent retrieval system can enhance knowledge access on outdated asset management platforms, highlighting the importance of intent understanding and semantic enrichment. Despite the small sample size, the results indicate significant potential for practical applications, warranting broader implementation and validation.