Semantic Layer Induction from Raw Telemetry via Hierarchical LLM and RAG Abstraction

RESUMO

Modern applications generate massive volumes of raw telemetry data, but translating those noisy, heterogeneous event streams into actionable business insights remains a fundamental challenge. Data engineers and analysts expend substantial effort reconciling semantic discrepancies, hand-crafting parsing logics, and maintaining fragile mappings between raw data and KPIs. In this paper, we present an end-to-end framework that fully automates the construction of semantic layer from application logs. Our approach introduces a two-stage abstraction: first, high-level features are identified via LLM inference augmented with domain-specific industry knowledge; second, fine-grained business nodes are derived through structured pipeline comprising data refinement, hybrid retrieval, multi-stage semantic filtering, clustering, and canonical naming. Evaluation on production-scale telemetry demonstrates our system improves human-assessed semantic quality from 50 to 80+ on 100-point scale, reduces maintenance effort by 80%, filters out 74% noise, and achieves 0.87 Cohen's kappa via integrated LLM-as-Judge evaluation, enabling continuous, scalable quality assurance. Overall, our work distinguishes itself from prior efforts by addressing the novel problem of semantic induction from raw telemetry, operating without labeled training data or manual rule engineering.