From Static Recommendations to Autonomous Decisions: A Framework for Agentic Recommendation Systems with Memory, Reasoning and Tool Use
DOI:
https://doi.org/10.56830/Keywords:
Agentic Recommendation Systems, Multi-Agent Recommender Systems, Memory-Augmented Recommendation, Tool-Using AI AgentsAbstract
Agentic recommendation systems can go beyond static rankings by grasping the intent behind actions, respecting constraints, calling for all necessary tools, communicating tradeoffs, and learning from feedback. However, controlled evidence on the cumulative contribution of these capabilities remains limited. This study created a five-agent system that includes a user-preference agent, product-retrieval agent, recommendation agent, explanation agent, and feedback-learning agent, coordinated by a deterministic orchestrator. The study conducted a comparative offline experiment using MovieLens 25M and Amazon reviews’23 electronics data, with 20,000 users, 20,000 ranking tasks, 300,000 sequential interactions, 10,000 feedback episodes, and 900 explanation judgments. The study measured popularity ranking, SASRec, single-agent LLM-RAG, and the proposed framework using NDCG@10, Recall@10, Hit Rate@10, MRR@10, constraint satisfaction, catalog coverage, diversity, novelty, explanation quality, reliability, latency, cost, and fallback behavior. The framework provided broader support and demonstrated traceability and auditability of memory, anchoring, reasoning, tool use, and recovery from failure. There was no model with an overall statistically significant advantage after Holm correction (adjusted p < .05). The available analytical outputs did not support reporting Krippendorff’s alpha, drift-recovery rate, incorrect-memory-update rate, post-feedback NDCG@10 changes, and operational threshold rates, which restricted claims of evaluator agreement, adaptation magnitude, and threshold compliance. The results show that absolute ranking accuracy is not the only metric for assessing an agentic recommendation system. Future studies should focus on live A/B testing at scale, dynamic catalogs, privacy-preserving memory, real feedback, fairness analysis for subgroups, and deployment costs.
References
A. Guo et al. (2024). “SoVAR: Building generalizable scenarios from accident reports for autonomous driving testing,” . in Proc. 39th IEEE/ACM Int. Conf. Automated Software Engineering, pp. 268–280. https://dl.acm.org/doi/pdf/10.1145/3691620.3695037.
A. Prasad. (2025). “Designing a reliable, ultra-low-latency data access environment for realtime applications in modern data centers,”. The USA Journals TAJIIR, vol. 7, no. 7, pp. 123–136, 2025. https://usajournal.net/index.php/tujtajiir/article/download/145/144.
A. Prasad. (2026). “Accelerating HPC performance: Strategies for overcoming modern challenges with NVMe, InfiniBand, and RDMA,”. in Proc. 2026 Int. Conf. Signal, Systems, and Computing for Next-Gen Automation (ICSSCNA), May 2026, pp. 463– 471. IEEE. https://ieeexplore.ieee.org/abstract/document/11546937/.
A. Ubale. (2023). “Beyond telematics: Leveraging generative AI for synthetic accident reconstruction and liability attribution in autonomous vehicle claims,”. Int. J. AI, BigData, Comput. Manage. Stud., vol. 4, no. 4, pp. 119–124, https://ijaibdcms.org/index.php/ijaibdcms/article/download/356/352.
A. Yehudai et al. (2025). “Survey on evaluation of LLM-based agents,” . arXiv:, 2503.16416.
https://arxiv.org/pdf/2503.16416.
Chapman, A., Lauro, L., Missier, P., & R. Torlone. (2024). “Supporting better insights of data science pipelines with fine-grained provenance,”. ACM Trans. Database Syst., vol. 49, no. 2, pp. 1–42, 2024. https://scholar.google.com/scholar?output=instlink&q=inf.
Chen, S., Piet, J., Sitawarin, C., & D. Wagner. (2025). “StruQ: Defending against prompt injection with structured queries,”. in Proc. 34th USENIX Security Symp. (USENIX Security 25), pp. 2383–2400. https://www.usenix.org/system/files/usenixsecurity25chen-sizhe.pdf.
Choudhary, S., & S. Sepuri. (2026). “AI-augmented decision governance in compliancecritical enterprise workflow systems: Implications for leadership and decision-making models in mid-sized enterprises,”. in Proc. 4th Int. Conf. Self Sustainable Artificial Intelligence Systems (ICSSAS), pp. 1183–1191. https://ieeexplore.ieee.org/abstract/document/11559573/.
Ding, P., & R. Stevens. (2025). “ToolRegistry: A protocol-agnostic tool management library for function-calling LLMs,”. arXiv:, 2507.10593, 2025. https://arxiv.org/pdf/2507.10593.
E. Tabassi. (2023). Artificial Intelligence Risk Management Framework (AI RMF 1.0). NIST AI 100-1. Gaithersburg, MD, USA: National Institute of Standards and Technology,.
Gannavarapu, P., & R. K. R. Samantapudi. (2025). “Secure AI-driven identity infrastructure for regulated sectors,”. Int. J. Intell. Syst. Appl. Eng., https://www.ijisae.org/index.php/IJISAE/article/view/7913.
Gondi, S., Sepuri, S., & P. Narayan. (2025). “AI ethics for public financial systems: A crosssector perspective,”. Int. J. Appl. Math, vol. 38, pp. 2212–2233.
Gupta, G., & R. Katarya. (2023). “A novel approach to alleviate data sparsity and generate dynamic fruit recommendations from point-of-sale data,”. Concurrency Comput. Pract. Exp., vol. 35, no. 1, Art. no. e7423, https://onlinelibrary.wiley.com/doi/abs/10.1002/cpe.7423.
H. Liao et al. (2026). “Eliminating out-of-domain recommendations in LLM-based recommender systems: A unified view,”. in Findings Assoc. Comput. Linguistics: ACL 2026, Jul. 2026, pp. 6251–6271. https://aclanthology.org/2026.findings-acl.310.pdf.
Heuss, M., Rijke, M. d., & A. Anand. (2025). “RankingSHAP: Faithful listwise feature attribution explanations for ranking models,”. in Proc. 48th Int. ACM SIGIR Conf. Research and Development in Information Retrieval, pp. 381–391. https://dl.acm.org/doi/pdf/10.1145/3726302.3729971 .
Huang, X., Lian, J., Lei, Y., Yao, J., Lian, D., & X. Xie. (2025). “Recommender AI agent:
Integrating large language models for interactive recommendations,” . ACM Trans. Inf. Syst., vol. 43, no. 4, pp. 1–33, 2025. https://dl.acm.org/doi/full/10.1145/3731446.
Ibrahim, O. A., Younis, E. M., Mohamed, E. A., & W. N. Ismail. (2025). “Revisiting recommender systems: An investigative survey,”. Neural Comput. Appl., vol. 37, no. 4, pp. 2145–2173. https://link.springer.com/content/pdf/10.1007/s00521-024-108285.pdf.
J. Zhang et al. (2024). “AgentCF: Collaborative learning with autonomous language agents for recommender systems,”. in Proc. ACM Web Conf., https://openreview.net/pdf?id=knAtuqBXDV.
Jannach, D., & H. Abdollahpouri. (2023). “A survey on multi-objective recommender systems,” . Frontiers Big Data, vol. 6, Art. no. 1157899, 2023. https://www.frontiersin.org/journals/bigdata/articles/10.3389/fdata.2023.1157899/pdf.
K. S. Chadha. (2025). “Machine learning-augmented ETL pipelines for fraud-resistant insurance claims processing,”. Int. J. Data Sci. Mach. Learn, vol. 5, no. 1, pp. 410– 436.
Kaul, A., Lan, Q., & P. Gupta. (2026). “AgentBound: Verifiable behavioral governance for autonomous AI agents,”. arXiv:, 30970, 2026. https://arxiv.org/pdf/2606.30970.
Khan, R., Joyce, D., & M. Habiba. (2025). “AGENTSAFE: A unified framework for ethical assurance and governance in agentic AI,”. arXiv:, 2512.03180, https://arxiv.org/pdf/2512.03180.
Kim, S., Moon, S., Tabrizi, R., Lee, N., Mahoney, M. W., Keutzer, K., et al. (2024). “An LLM compiler for parallel function calling,”. in Proc. 41st Int. Conf. Machine Learning, Jun. 2024. , https://openreview.net/pdf?id=uQ2FUoFjnF.
Liao, J., Yang, M., Zhou, W., Zhang, H., & J. Wen. (2024). “Modeling item exposure and user satisfaction for debiased recommendation with causal inference,”. Inf. Sci., vol. 676, Art. no. 120834, https://www.sciencedirect.com/science/article/pii/S0020025524007485 .
Liao, M., & S. S. Sundar, J. B. (2022). “User trust in recommendation systems: A comparison of content-based, collaborative and demographic filtering,”. in Proc. 2022 CHI Conf.
Human Factors Comput, Syst., Apr. 2022, pp. 1–14. https://dl.acm.org/doi/pdf/10.1145/3491102.3501936.
Lin, S., Yang, L., Fan, D., & J. Zhang. (2022). “Beyond not-forgetting: Continual learning with backward knowledge transfer,”. in Advances in Neural Information Processing
Systems, vol. 35, pp. 16165–16177,
https://proceedings.neurips.cc/paper_files/paper/2022/file/6728fcf94660c59c938319a6 833a6073-Paper-Conference.pdf.
Lyu, Y., Cao, J., Xiao, Y., Sun, Z., Niu, B., Li, F., et al. (2026). “Privacy risks of intermediate representations: Attribute inference in distributed LLM inference,”. in Findings Assoc. Comput. Linguistics: ACL 2026,, pp. 23412–23426. https://aclanthology.org/2026.findings-acl.1172.pdf.
M. Ravaut, H. Z. (2024). “Parameter-efficient conversational recommender system as a language processing task,”. in Proc. 18th Conf. Eur. Chapter Assoc. Comput. Linguistics, vol. 1: Long Papers, Mar. 2024, pp. 152–165. https://aclanthology.org/2024.eacl-long.9.pdf.
M. T. Tran. (2026). “Advancing retail and service strategies: AI-driven consumer behavior prediction, gamification, and ethical marketing,”. J. Retailing Consum. Serv., vol. 88,
Art. no. 104558, 2026. https://magisterenmarketing.cl/wp-
content/uploads/2026/01/Advancing-retail-and-service-strategies-AI-drivenconsumer-behavior.pdf.
P. Gannavarapu. (2025). “Deploying Azure AD federation with SAML for secure enterprise SaaS integration,”. Computer Fraud & Security, https://computerfraudsecurity.com/index.php/journal/article/view/782.
P. K. R. Prakashkumar. ((n.d)). “Secure bank integration framework for Oracle ERP Fusion: Enhancing payment disbursement, auto lockbox, and bank account reconciliation,”. https://www.eelet.org.uk/index.php/journal/article/view/4082.
Pamuk, M., Schumann, M., & R. C. Nickerson. (2024). “What do the regulators mean? A taxonomy of regulatory principles for the use of AI in financial services,” . Mach. Learn. Knowl. Extr., vol. 6, no. 1, pp. 143–155, 2024. https://www.mdpi.com/25044990/6/1/8.
Patil, S. G., Mao, H., Yan, F., Ji, C.-J. C., Suresh, V., Stoica, I., et al. (2025). “The Berkeley Function Calling Leaderboard (BFCL): From tool use to agentic evaluation of large language models,”. in Proc. 42nd Int. Conf. Machine Learning,, https://openreview.net/pdf?id=2GmDdhBdDk.
Peng, Q., Liu, H., Huang, H., Yang, Q., & M. Shao. (2025). “A survey on LLM-powered agents for recommender systems,” . arXiv, 2502.10050. https://aclanthology.org/anthology-files/pdf/findings/2025.findings-emnlp.620.pdf.
Polleti, G. P., Souza, D. L., & F. G. Cozman. (2025). “Generating explanations for knowledge-aware conversational recommendation systems,”. User Model. User-
Adapt. Interact, vol. 35, no. 3, Art. no. 12, https://link.springer.com/content/pdf/10.1007/s11257-025-09432-6.pdf.
Rahmani, H. A., Deldjoo, Y., Tourani, A., & M. Naghiaei. (2022). “The unfairness of active users and popularity bias in point-of-interest recommendation,”. in Advances in Bias and Fairness in Information Retrieval: Proc. 3rd Int. Workshop BIAS 2022, pp. 56– 68. Cham, Switzerland: Springer. https://arxiv.org/pdf/2202.13307.
Raj, S., Jha, A. V., Anand, M., & S. Saha. (2026). “HARPO: Hierarchical agentic reasoning for user-aligned conversational recommendation,”. in Proc. 64th Annu. Meeting Assoc. Comput. Linguistics, vol. 1: Long Papers, Jul. 2026, pp. 35580–35599.. https://aclanthology.org/2026.acl-long.1646.pdf.
Rangu, S., Chadha, K. S., & S. Samala. (2025). “AI-powered cloud architecture, agile analytics, and enterprise DevOps-enabled framework for healthcare and insurance ecosystem optimisation,”. in Proc. 10th Int. Conf. Science Technology Engineering and Mathematics (ICONSTEM), Chennai, India, pp. 1–7. https://ieeexplore.ieee.org/abstract/document/11374414/.
S. Durgam. (2025). “Peer benchmarking systems for RIA performance evaluation in investment technology,”. https://computerfraudsecurity.com/index.php/journal/article/view/785.
S. K. RaoVanama. (2024). “AI-augmented CI/CD pipeline optimization for scalable cloudnative deployment,”. Int. J. Artif. Intell. Data Sci. Mach. Learn., vol. 5, no. 4, pp. 175– 187, 2024. https://ijaidsml.org/index.php/ijaidsml/article/download/368/338.
Sharma, S., Dubey, G. P., & H. K. Shakya. (2024). “Optimizing user satisfaction in movie recommendations using variable learning rates and dynamic neighborhood functions in SOMs,”. Int. J. Exp. Res. Rev., vol. 41, pp. 130–145, 2024. https://www.academia.edu/download/117189744/11._Optimizing_User_Satisfaction_i n_Movie_Recommendations_Using_Variable_Learning_Rates_and_Dynamic_Neighb orhood.pdf.
Shrimal, A., Kanagaraj, S., Biswas, K., Raghuraman, S., Nediyanchath, A., Zhang, Y., et al. (2024). “MARCO: Multi-agent real-time chat orchestration,”. in Proc. 2024 Conf. Empirical Methods Natural Language Process.: Industry Track, Nov. 2024, pp. 1381– 1392. https://aclanthology.org/2024.emnlp-industry.102.pdf.
T. Schick et al. (2023). “Toolformer: Language models can teach themselves to use tools,” .
in Advances in Neural Information Processing Systems, vol. 36, pp. 68539–68551, 2023.
https://proceedings.neurips.cc/paper_files/paper/2023/file/d842425e4bf79ba039352da 0f65.
W. Zhang et al. (2026). “PersonaAgent: Bridging memory and action for personalized LLM agents,”. in Findings Assoc. Comput. Linguistics: ACL 2026, Jul. , pp. 26421–26439.
https://aclanthology.org/2026.findings-acl.1315.pdf.
X. Shen et al. (2026). “Agentic recommender system with hierarchical belief-state memory,” . arXiv, 2605.14401, 2026. https://arxiv.org/pdf/2605.14401.
Y. Benajiba et al. (2025). “Open Agent Specification (Agent Spec) technical report,”. arXiv:, 2510.04173, https://ui.adsabs.harvard.edu/abs/2025arXiv251004173B/abstract.
Y. Shang et al. (2026). “AgentRecBench: Benchmarking LLM agent-based personalized recommender systems,”. in Advances in Neural Information Processing Systems, vol.
38, https://proceedings.neurips.cc/paper_files/paper/2025/file/e2d6f7249add096e26679ea de.
Y. Wang et al. (2026). “From agent traces to trust: A survey of evidence tracing and execution provenance in LLM agents,”. arXiv:, 2606.04990. https://arxiv.org/pdf/2606.04990.
Yoon, S.-E., He, Z., Echterhoff, J. M., & J. McAuley. (2024). “Evaluating large language models as generative user simulators for conversational recommendation,” . arXiv:, 2403.09738, https://arxiv.org/pdf/2403.09738.
Z. Wu et al. (2026). “AgentGraph: Trace-to-graph platform for interactive analysis and robustness testing in agentic AI systems,”. in Proc. AAAI Conf. Artif. Intell, vol. 40, no. 48, Mar. 2026, pp. 41721–41723. hhttps://ojs.aaai.org/index.php/AAAI/article/download/42393/46354.
Zhou, H., & H. Y. Chan. (2026). “ORCH: Many analyses, one merge—A deterministic multiagent orchestrator for discrete-choice reasoning with EMA-guided routing,”. arXiv, 2602.01797. https://arxiv.org/pdf/2602.01797.
Zhu, J., Dai, Q., Su, L., Ma, R., Liu, J., & G. Cai. (2022). “BARS: Towards open benchmarking for recommender systems,”. in Proc. 45th Int. ACM SIGIR Conf. Research and Development in Information Retrieval, Jul. 2022, pp. 2912–2923.
