RESEARCH INTO THE APPLICATION OF VECTOR DATABASES IN GENERATIVE ARTIFICIAL INTELLIGENCE
DOI:
https://doi.org/10.28925/2663-4023.2026.33.1248Keywords:
databases, artificial intelligence, vector databases (VDB), generative artificial intelligence (GenAI), large language models (LLM), Retrieval-Augmented Generation (RAG), semantic search, vector embeddings, HNSW, RAFT, scalability, cognitive infrastructure, ANN, quantizationAbstract
This paper studies the use of vector databases in generative artificial intelligence. The purpose of this article is to study the possibility of using vector databases as the foundation of modern AI infrastructure and their role in expanding the cognitive capabilities of GenAI. The object of the study is the process of using databases in artificial intelligence. The subject of the study is the use of vector databases in generative artificial intelligence. The following tasks were solved in this study: the role of vector repositories in expanding the cognitive capabilities of generative AI was investigated; a comparative analysis of traditional relational systems and vector databases was carried out; the mechanics and verification potential of the RAG and RAFT architectural paradigms were analyzed. Prospects for further research are identified, which consist of the following: development of effective mechanisms for monitoring and automatic correction of semantic drift of embeddings in conditions of constant updating of corporate knowledge; research into adaptive memory management strategies to eliminate threshold performance effects when scaling vector indexes; research into the possibilities of combining hybrid search architectures that combine semantic and full-text indexing in a single technological environment; standardization of metrics for assessing the quality of vector search in the context of specific requirements of industrial RAG systems, which will ensure a more informed choice of infrastructure solutions for specific application tasks.
Downloads
References
FAISS. (n.d.). Documentation. https://faiss.ai/index.html
Taipalus, T. (2024). Vector database management systems: Fundamental concepts, use-cases, and current challenges. Cognitive Systems Research, 85. https://doi.org/10.1016/j.cogsys.2024.101216
Zhang, T., et al. (2024). RAFT: Adapting language model to domain specific RAG. arXiv. https://doi.org/10.48550/arXiv.2403.10131
Gao, L., et al. (2023). Retrieval-augmented generation for large language models: A survey. arXiv. https://doi.org/10.48550/arXiv.2312.10997
Pan, X., et al. (2023). Survey of vector database management systems. arXiv. https://doi.org/10.48550/arXiv.2310.14021
Shi, Y., et al. (2024). Enhancing retrieval and managing retrieval: A four-module synergy for improved quality and efficiency in RAG systems. arXiv. https://doi.org/10.48550/arXiv.2407.10670
Ma, L., et al. (2023). A comprehensive survey on vector database: Storage and retrieval technique, challenge. arXiv. https://doi.org/10.48550/arXiv.2310.11703
Wang, M., et al. (2023). Embedding in recommender systems: A survey. arXiv. https://doi.org/10.48550/arXiv.2310.18608
Milvus. (n.d.). Vector database documentation. https://milvus.io
Pinecone. (n.d.). Vector database guide. https://www.pinecone.io/learn/vector-database/
Gao, Y., et al. (2023). Retrieval-augmented generation for large language models: A survey. arXiv. https://doi.org/10.48550/arXiv.2312.10997
Huang, L., et al. (2023). A survey on hallucination in large language models: Principles, taxonomy, challenges, and open questions. arXiv. https://doi.org/10.48550/arXiv.2311.05232
Gupta, S., Ranjan, R., & Singh, S. N. (2024). A comprehensive survey of retrieval-augmented generation (RAG): Evolution, current landscape and future directions. arXiv. https://doi.org/10.48550/arXiv.2410.12837
Pan, J. J., Wang, J., & Li, G. (2024). Survey of vector database management systems. https://dbgroup.cs.tsinghua.edu.cn/ligl/papers/vldbj2024-vectordb.pdf
Schwaber-Cohen, R. (n.d.). What is a vector database & how does it work? Use cases + examples. Pinecone. https://www.pinecone.io/learn/vector-database/
Tu, Y., Su, W., Zhou, Y., Liu, Y., & Ai, Q. (2025). RbFT: Robust fine-tuning for retrieval-augmented generation against retrieval defects. arXiv. https://doi.org/10.48550/arXiv.2501.18365
Gu, J. (2024). A research of challenges and solutions in retrieval augmented generation (RAG) systems. Humanities and Social Sciences Communications / DRPress (HSET). https://doi.org/10.54097/364hex16
Wallace, J. A. (n.d.). What are the most common vector database challenges? Redis. https://redis.io/blog/common-challenges-working-with-vector-databases/
Kang, D., Jiang, D., Yang, H., Liu, H., & Li, B. (2025). Scalable disk-based approximate nearest neighbor search with page-aligned graph. arXiv. https://doi.org/10.48550/arXiv.2509.25487
Subramanya, S. J., Devvrit, Kadekodi, R., Krishaswamy, R., & Simhadri, H. V. (2019). DiskANN: Fast accurate billion-point nearest neighbor search on a single node. In Proceedings of the 33rd International Conference on Neural Information Processing Systems (pp. 13766-13776). Curran Associates. https://dl.acm.org/doi/abs/10.5555/3454287.3455520
Chen, Q., Zhao, B., Wang, H., Li, M., Liu, C., Li, Z., Yang, M., & Wang, J. (2021). SPANN: Highly-efficient billion-scale approximate nearest neighbor search. arXiv. https://doi.org/10.48550/arXiv.2111.08566
Microsoft Research. (n.d.). DiskANN: Vector search for all. https://www.microsoft.com/en-us/research/project/project-akupara-approximate-nearest-neighbor-search/
Jiang, P., Ouyang, S., Jiao, Y., Zhong, M., Tian, R., & Han, J. (2025). A survey on retrieval and structuring augmented generation with large language models. arXiv. https://arxiv.org/html/2509.10697v1
Vejendla, H. (2025). Drift-adapter: A practical approach to near zero-downtime embedding model upgrades in vector databases. arXiv. https://arxiv.org/pdf/2509.23471
Milvus. (n.d.). RRF ranker. https://milvus.io/docs/rrf-ranker.md
Li, Z., Wang, Z., Wang, W., Hung, K., Xie, H., & Wang, F. (2025). Retrieval-augmented generation for educational application: A systematic survey. Computers and Education: Artificial Intelligence, 8, 100417. https://doi.org/10.1016/j.caeai.2025.100417
Brown, A., Roman, M., & Devereux, B. (2025). A systematic literature review of retrieval-augmented generation: Techniques, metrics, and challenges. arXiv. https://doi.org/10.48550/arXiv.2508.06401
Johnson, J., Douze, M., & Jégou, H. (2017). Billion-scale similarity search with GPUs. arXiv. https://doi.org/10.48550/arXiv.1702.08734
Chakraborty, A., Dahal, C., & Gupta, V. (2025). Federated retrieval-augmented generation: A systematic mapping study. arXiv. https://arxiv.org/pdf/2505.18906
Amazon Web Services. (n.d.). Cost considerations for vector databases. https://docs.aws.amazon.com/prescriptive-guidance/latest/choosing-an-aws-vector-database-for-rag-use-cases/cost.html
Microsoft Azure. (n.d.). Pricing for search services. https://azure.microsoft.com/en-us/pricing/details/search/
TensorBlue Blog. (2025). Best vector database 2025: Pinecone vs Weaviate vs Qdrant vs Milvus. https://tensorblue.com/blog/vector-database-comparison-pinecone-weaviate-qdrant-milvus-2025
Theodo. (n.d.). How to choose your vector database. https://www.theodo.com/en-fr/blog/how-to-choose-your-vector-database
Pinecone. (n.d.). Official website. https://www.pinecone.io/
Milvus. (n.d.). Official website. https://milvus.io/
Weaviate. (n.d.). Official website. https://weaviate.io/
Qdrant. (n.d.). Official website. https://qdrant.tech/
pgvector. (n.d.). GitHub repository. GitHub. https://github.com/pgvector/pgvector
Amazon Web Services. (n.d.). What is Elasticsearch? https://aws.amazon.com/what-is/elasticsearch/
Kuznetsov, O., Smirnov, O., Akhmetov, B., Alimseitova, Z., & Imoize, A. L. (2025). Deep learning frontiers in copy-move forgery detection: Advances, challenges, and future directions. In Advancements in Cybersecurity Next Generation Systems and Applications (pp. 202-229). https://doi.org/10.1201/9781003546153
Smirnov, O., Fedorov, E., Neskorodieva, A., & Neskorodieva, T. (2024). Intellectual classification method of gymnastic elements based on combinations of descriptive and generative approaches. CEUR Workshop Proceedings, 3664, 11-23.
Al-Mudhafar, A. A., Smirnova, T., Buravchenko, K., & Smirnov, O. (2023). The method of assessing and improving the user experience of subscribers in software-configured networks based on the use of machine learning. Advanced Information Systems, 7(2), 49-56.
Smirnov, O., Karapetyan, A., & Fedorov, E. (2022). Creating neural network and single solution human-based metaheuristic methods of solving the traveling salesman problem. CEUR Workshop Proceedings, 3312, 47-58.
Lysenko, I., Mynailenko, R., Smirnov, S., Buravchenko, K., Yakymenko, N., & Smirnov, O. (2025). Research of artificial intelligence tools for intelligent data analysis. Cybersecurity: Education, Science, Technique, 3(31), 227-241. https://doi.org/10.28925/2663-4023.2025.31.1022
Usik, P. S., Smirnova, T. V., Buravchenko, K. O., Smirnov, O. A., Ulichev, O. S., & Smirnov, S. A. (2025). Research of cybersecurity technologies for banking systems using artificial intelligence. Cybersecurity: Education, Science, Technique, 1(29), 704-716. https://doi.org/10.28925/2663-4023.2025.29.930
Smirnov, O. A., Konstantynova, L. V., Konoplitska-Slobodeniuk, O. K., Kozirova, N. V., Yakymenko, N. M., Dorenskyi, O. P., & Buravchenko, K. O. (2025). Research of artificial intelligence tools for working with databases and data analysis. Cybersecurity: Education, Science, Technique, 3(27), 429-448. https://doi.org/10.28925/2663-4023.2025.27.763
Al-Mudhafar, A. A. A., Smirnova, T. V., Buravchenko, K. O., & Smirnov, O. A. (2023). Method of assessment and improvement of subscriber user experience in software-defined networks based on machine learning. Advanced Information Systems, 7(2), 49-56. https://doi.org/10.20998/2522-9052.2023.2.07
Published
How to Cite
Issue
Section
License
Copyright (c) 2026 Олексій Смірнов, Роман Ткачук, Наталія Козірова, Лілія Константинова, Оксана Конопліцька-Слободенюк, Наталія Якименко, Сергій Смірнов
This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.
