LLM-powered RAG Earnings Call Analyzer

International Journal of Emerging Research in Science, Engineering, and Management
Vol. 2, Issue 4, pp. 114-118, April 2026.

https://doi.org/10.58482/ijersem.v2i4.16

This work is licensed under a Creative Commons Attribution 4.0 International License.

LLM-powered RAG Earnings Call Analyzer

K G Mohanavalli, Golla Adi, D R Mounika, P Dilli Prasad, C Karthikeya Anjan Kumar, D Dhanush

Department of CSE, Siddartha Institute of Science and Technology, Puttur, India.

Full Text PDF

Abstract: : Analysis of earnings call transcripts plays a critical role in understanding corporate performance, managerial intent, and future financial outlook. However, the large volume and unstructured nature of financial transcripts make manual interpretation time-consuming and inefficient. Recent advances in Natural Language Processing and Large Language Models have enabled automated extraction of meaningful insights from financial communication data. Retrieval-Augmented Generation integrates semantic retrieval with contextual language modeling to improve factual consistency and analytical depth during transcript processing. Hybrid indexing techniques combining lexical search and dense vector embeddings support accurate context selection across multiple documents. Transformer-based architectures further enhance sentiment detection, speaker attribution, key metric identification, and structured summarization of financial discussions. Multi-document retrieval enables comparative trend analysis across reporting periods, supporting improved decision-making efficiency. The integration of retrieval mechanisms with generative language models provides scalable, context-aware, and reliable analysis of earnings call transcripts for financial intelligence applications.

Keywords: Retrieval-Augmented Generation, Earnings Call Analysis, Financial Text Summarization, Sentiment Analysis, Vector Embeddings.

References: 

  1. G B. V. Nagendra, S. Kumar Chandar, J. B. Simha and J. A. John Bazil, “Financial Lexicon based Sentiment Prediction for Earnings Call Transcripts for Market Intelligence,” 2024 5th International Conference on Image Processing and Capsule Networks (ICIPCN), Dhulikhel, Nepal, 2024, pp. 595-603, doi: 10.1109/ICIPCN63822.2024.00103.
  2. N. Bv, J. B. Simha, S. Abhi and K. C. S, “Deploying NLP Techniques for Earnings Call Transcripts for Financial Analysis: A Reverse Phenomenon Paradigm,” 2023 7th International Conference on I-SMAC (IoT in Social, Mobile, Analytics and Cloud) (I-SMAC), Kirtipur, Nepal, 2023, pp. 368-375, doi: 10.1109/I-SMAC58438.2023.10290494.
  3. Y. Huang, W. Tai, F. Zhou, Q. Gao, and T. Zhong, “Extracting key insights from earnings call transcript via information-theoretic contrastive learning,” Information Processing & Management, vol. 62, no. 3, p. 103998, Nov. 2024, doi: 10.1016/j.ipm.2024.103998.
  4. T. Xu, Z. Piao, T. Mukai, Y. Murayama, and K. Izumi, “Can large language models autonomously generate unique and profound insights in fundamental analysis?,” Intelligent Systems With Applications, vol. 27, p. 200566, Aug. 2025, doi: 10.1016/j.iswa.2025.200566.
  5. A. Huang, H. Wang, and Y. Yang, “FinBERT—A deep learning approach to extracting textual information,” SSRN Electronic Journal, Jan. 2020, doi: 10.2139/ssrn.3910214.
  6. V. S. Nageen, “NLP-Driven Sentiment Analysis of earnings calls and its impact on stock volatility,” International Journal of AI BigData Computational and Management Studies, vol. 5, no. 4, pp. 176–181, Dec. 2024, doi: 10.63282/3050-9416.ijaibdcms-v5i4p119.
  7. Y. Chen, H. Dai, W. Li, F. Wang, B. Wang and L. Wang, “Time Reversal Enabled Fiber-Optic Time Synchronization,” in IEEE Transactions on Instrumentation and Measurement, vol. 72, pp. 1-8, 2023, Art no. 5503508, doi: 10.1109/TIM.2023.3300428.
  8. Y. Sun, H. Yuan, and F. Xu, “Financial sentiment analysis for pre-trained language models incorporating dictionary knowledge and neutral features,” Natural Language Processing Journal, vol. 11, p. 100148, Apr. 2025, doi: 10.1016/j.nlp.2025.100148.
  9. A. Todd, J. Bowden, M. Cummins, and Y. Su, “A multimodal sentiment classifier for financial decision making,” International Review of Financial Analysis, vol. 105, p. 104322, May 2025, doi: 10.1016/j.irfa.2025.104322.
  10. H. Lee, J. H. Kim, and H. S. Jung, “From corporate earnings calls to social impact: Exploring ESG signals in S&P 500 ESG index companies through transformer-based models,” Journal of Cleaner Production, vol. 501, p. 145320, Mar. 2025, doi: 10.1016/j.jclepro.2025.145320.
  11. M. Leippold, “Sentiment spin: Attacking financial sentiment with GPT-3,” Finance Research Letters, vol. 55, p. 103957, May 2023, doi: 10.1016/j.frl.2023.103957.
  12. R. Abilio, G. P. Coelho, and A. E. A. Da Silva, “Evaluating Named Entity Recognition: A comparative analysis of mono- and multilingual transformer models on a novel Brazilian corporate earnings call transcripts dataset,” Applied Soft Computing, vol. 166, p. 112158, Aug. 2024, doi: 10.1016/j.asoc.2024.112158.
Download PDF

This article is part of Volume 2, Issue 4 (April 2026)