Component obsolescence and supply-chain disruptions increasingly force engineers to spend significant time manually searching and comparing PDF datasheets to identify compatible replacement parts. We propose an AI-powered datasheet assistant based on a Retrieval-Augmented Generation (RAG) pipeline that automatically processes datasheets to accelerate component identification and matching. The core contribution is a summary-driven retrieval mechanism: a Large Language Model (LLM) generates a structured semantic summary of an input datasheet, and the vector embedding of this summary is used to retrieve semantically similar components from a reference database. The system also supports natural language question answering and structured component comparison. Its architecture separates scalable text-only reference indexing from more expensive query-time summarization and reranking. Validation includes a controlled synthetic benchmark and a pilot-scale real-world evaluation on 18 publicly listed microcontroller datasheets grouped into six engineering families. The synthetic benchmark is used to assess pipeline behavior under controlled conditions, while the real-world evaluation measures performance on heterogeneous manufacturer datasheets. In the real-world evaluation, structured summaries generated with Claude Sonnet 4.5 combined with cross-encoder reranking achieved a 72.2% Family Retrieval Rate at $k=1$ (13/18; Wilson 95% CI: 49.1–87.5%). Additional experiments with local LLM summaries indicate that retrieval performance depends strongly on summary quality and model capability, with lightweight local summarizers producing lower first-candidate retrieval performance in this setup. The analysis further reports confidence intervals, no-summary baselines, chunking sensitivity, and an Image Reference Rate metric used as a lexical reference proxy rather than a direct measure of visual grounding.

Background/Objectives: The COVID-19 pandemic exposed the need for economic resilience in dental practices, which faced reduced patient numbers, increased costs, and service restrictions. This study aimed to develop a predictive model for dental practice resilience and to identify the financial characteristics associated with better performance during and after a crisis. Methods: We analyzed official financial statements for 2,474 Bucharest dental practices. After applying mode-specific inclusion criteria, the final analytic cohorts comprised 850 firms for crisis-year prediction in 2020 and 801 firms for the three 2021 recovery scenarios. The Dental Practice Resilience Index (DPRI) was developed as a composite measure of profitability, expense efficiency, debt-to-turnover, turnover per employee, and profit per employee. Seven regression model families were evaluated across four prediction scenarios, and continuous predictions were also translated into tertile-based risk classes for decision support. Results: Crisis-year prediction showed moderate but useful accuracy, with Linear Regression narrowly achieving the highest cross-validated performance ( R ² = 0.537 ± 0.078). Predicting 2021 recovery from pre-crisis data alone was more difficult, and Random Forest narrowly ranked first ( R ² = 0.452 ± 0.084). The strongest scenario-level results were obtained when observed 2020 data were included, with Elastic Net reaching R ² = 0.643 ± 0.071 and Random Forest reaching balanced accuracy = 0.648 ± 0.044. When observed crisis-year data were unavailable, a two-step synthetic recovery approach recovered part of that information, but remained below the crisis-informed configuration. Conclusions: Machine learning methods proved effective not only in predicting dental practice resilience during a crisis, but also in anticipating post-crisis recovery potential. The DPRI provided an interpretable resilience measure, while the combination of continuous prediction and tertile-based classification offered both nuanced forecasts and practical risk stratification. Observed crisis-year information substantially improved recovery forecasting, although pre-crisis indicators alone still retained meaningful predictive value.