Employing Approaches of Raising Quality of Generated Answers of AI LLM-Systems when Accumulated Enterprise Information is Used

The development of large language models (LLM) has extended opportunities of processing non-structured data in corporate environment. However, LLM fundamental restriction lies in their dependence on data of preliminary learning, which can decrease their ability to work safely with enterprise information non-published in other sources. The article provides analysis of current approach to using corporate data to upgrade quality of answers generated by LLM. Systems of Retrieval-Augmented Generation (RAG) are studied both in classical vector realization and in GraphRAG. Agent systems and integration of RAGand GraphRAG approaches in their architecture were discussed, as it can make it possible to build complicated systems of AI and solutions requiring multi-stage analysis of corporate non-structured data. Architecture of classical vector RAG-approach is described, which consists of three key stages: getting vector presentation of indexed fragments of papers; searching for fragments relevant to user requirement; forming the answer on extracted context combined with the initial request. As a key restriction the author showed the absence of mechanism necessary for integral analysis of loaded proprietary data. In its turn GraphRAG uses mechanism of plotting knowledge graph on context, which can help work not with semantically similar data fragments but with hierarchically clustered information by accumulated analysis of text summaries. Shortcomings and benefits of using approaches were formulated, criteria of expediency of RAG and GraphRAG practical application were identified and examples of systems for their use were provided

1. Winer N. Kibernetika, ili Upravlenie i svyaz v zhivotnom i mashine [Cybernetics or Management and Communication in Living-Being and Machine], translated from English by I. V. Solovev and G. N. Povarov; edited by G. N. Povarov. 2nd edition. Moscow, ‘Nauka’ the General Editors for Foreign Countries, 1983. (In Russ.).

2. Denisenko V. Yu. Monitoring effektov tsifrovykh produktov v usloviyakh tsifrovoy transformatsii promyshlennykh predpriyatiy [Monitoring Effects of Digital Products in Conditions of Digital Transformation in Industrial Enterprises]. Kreativnaya ekonomika [Creative Economy], 2021, Vol. 15, No. 5, pp. 1715–1724. (In Russ.).

3. Kokuytseva T. V., Ovchinnikova O. P. Metodicheskie podkhody k otsenke effektivnosti tsifrovoy transformatsii predpriyatiy vysokotekhnologichnykh otrasley promyshlennosti [Methodological Approaches to Estimating Efficiency of Digital Transformation in Enterprises of Highly-Technological Industries]. Kreativnaya ekonomika [Creative Economy], 2021, Vol. 15, No. 6, pp. 2413–2430. (In Russ.).

4. Kochetkov E. P., Zabavina A. A., Gafarov M. G. Tsifrovaya transformatsiya kompaniy kak instrument antikrizisnogo upravleniya: empiricheskaya otsenka vliyaniya na effektivnost [Digital Transformation of Companies as Tool of Crisis Management: Empiric Assessment of Impact on Efficiency]. Strategicheskie resheniya i risk-menedzhment [Strategic Solutions and RiskManagement], 2021, Vol. 12, No. 1, pp. 68–81. (In Russ.).

5. Kurenkov A. L. Upravlenie tsifrovoy transformatsiey kommercheskikh predpriyatiy: monografiya [Managing Digital Transformation of Commercial Enterprises: monograph]. Moscow, Infra-M, 2025. (In Russ.).

6. Ukolov V. F., Afanasev V. Ya., Cherkasov V. V. Klyuchevye effekty tsifrovizatsii i vozmozhnye poteri [Key Effects of Digitalization and Possible Losses]. Vestnik universiteta [Bulletin of the University], 2019, No. 8, pp. 55–58. (In Russ.).

7. Tsenzharik M. K., Krylova Yu. V., Steshenko V. I. Tsifrovaya transformatsiya kompaniy: strategicheskiy analiz, faktory vliyaniya i modeli [Digital Transformation of Companies: Strategic Analysis, Impact Factors and Models]. Vestnik Sankt-Peterburgskogo universiteta. Ekonomika [Bulletin of the St. Petersburg University. Economics], 2020, Vol. 36, Issue 3, pp. 390–420. (In Russ.).

8. Brown T., Mann B., Ryder N. et al. Language Models are Few-Shot Learners. Advances in Neural Information Processing Systems , 2020, Vol. 33, pp. 1877–1901.

9. Edge D., Trinh H., Cheng N. et al. From Local to Global: A GraphRAG Approach to Query-Focused Summarization. Available at: https://arxiv.org/abs/2404.16130 (accessed 19.02.2026).

10. Fortunato S. Community Detection in Graphs. Physics Reports , 2010, Vol. 486, No. 35, pp. 75–174.

11. Jeffcock P. What's the Difference between AI, Machine Learning, and Deep Learning? Peter Jeffcock. Available at: https://blogs.oracle.com/bigdata/difference-ai-machine-learningdeep-learning (accessed 09.03.2020).

12. Lewis P., Perez E., Piktus A. et al. Retrieval-Augmented Generation for Knowledge Intensive NLP Tasks. Advances in Neural Information Processing Systems , 2020, Vol. 33, pp. 9459–9474.

13. Liu N. F., Lin K., Hewitt J. et al. Lost in the Middle: How Language Models Use Long Contexts. Available at: https://arxiv.org/abs/2307.03172 (accessed 19.02.2026).

14. Shinn N., Cassano F., Gopinath A. et al. Reflexion: Language Agents with Verbal Reinforcement Learning. Available at: https://arxiv.org/abs/2303.11366 (accessed 19.02.2026).

15. Traag V. A., Waltman L., Van Eck N. J. From Louvain to Leiden: Guaranteeing Well Connected Communities. Scientific Reports , 2019, Vol. 9, No. 1, Article 5233.

16. Vaswani A., Shazeer N., Parmar N. et al. Attention is All You Need. Available at: https://arxiv.org/abs/1706.03762 (accessed 19.02.2026).

17. Wei J., Wang X., Schuurmans D. et al. Chain-of-Thought Prompting Elicits Reasoning in Large Language Models. Available at: https://arxiv.org/abs/2201.11903 (accessed 19.02.2026).

18. Yao S., Zhao J., Yu D. et al. ReAct: Synergizing Reasoning and Acting in Language Models // International Conference on Learning Representations (ICLR), 2023. Available at: https://collaborate.princeton.edu/en/publications/react-synergizing-reasoning-and-actingin-language-models/ (accessed 19.02.2026).