RAG-BASED AUTOMATION OF THE CLIENT JOURNEY IN MEDICAL AND WELLNESS SYSTEMS: OPERATIONAL EFFICIENCY, CLIENT RETENTION, AND BEHAVIORAL CALIBRATION OF AI-MEDIATED COMMUNICATION

Aponchuk Alona

doi:10.69635/mssl.2026.2.2.45

Authors

Aponchuk Alona M.Sc. in Social Informatics, QA Engineer, Specialist in CRM, Patient Management and Digital Workflow Systems, United States Author https://orcid.org/0009-0008-3505-7871

DOI:

https://doi.org/10.69635/mssl.2026.2.2.45

Keywords:

RAG, CRM, Wellness Systems, Medical Client Interaction, CAC, Client Retention, Chatbot Automation, PersonaMatrix, Behavioral Calibration, Human–AI Interaction

Abstract

This article examines Retrieval-Augmented Generation (RAG)-based automation of client interaction in medical and wellness systems as a scalable operational model for reducing customer acquisition cost (CAC), improving client retention, and increasing the reliability of AI-mediated service communication. The study introduces a structured operational framework combining acquisition funnel analysis, retention funnel analysis, time-cost evaluation, staff workload assessment, and relative efficiency metrics.

The model demonstrates that CAC is determined not only by advertising expenditure but also by fragmented interaction processes across the client journey: initial response, consultation, follow-up, appointment coordination, and post-service retention. Based on a modeled funnel of 10,000 impressions, 200 clicks, 150 leads, 90 consultations, and 60 clients , a manual interaction scenario requires 40 minutes per client, resulting in 40 hours of total staff work for 60 clients and a CAC of approximately $83. Following the implementation of an integrated automation architecture uniting CRM data, verified service knowledge, and RAG-driven chatbot interactions, processing time decreases to 15 minutes per client, staff workload drops from 6.6 to 2.5 hours per employee, and CAC decreases to approximately $66.

Furthermore, the article proposes a behavioral calibration model for RAG-based systems. Drawing on measurement methodologies implemented in the PersonaMatrix project (Opulentia SC LLC, Florida, USA), the study argues that RAG systems in health-adjacent contexts must be evaluated by response stability, interpretive drift, and structural coherence. Three indicators—Response Stability Index (RSI), Interpretive Drift Score (IDS), and Response Coherence/Structure Score (RCS)—are proposed as a calibration layer to transform automated interfaces into context-sensitive, ethically bounded service mediators.

References

Amugongo, L. M., et al. (2024). Retrieval augmented generation for large language models in healthcare: A systematic review. Journal of Medical Artificial Intelligence, 7, 12–25.

Aponchuk, A. (2026). RAG-based automation of the client journey in medical and wellness systems: Operational efficiency, client retention, and behavioral calibration of AI-mediated communication. Journal of Digital Workflow and Patient Management, 4(1), 89–104.

Awotunde, O. J. (2025). Continuous model calibration. Artificial Intelligence Review, 58(3), 201–215. https://doi.org/10.1007/s10462-025-10142-9

Babu, A., & Boddu, S. (2024). BERT-based medical chatbot. International Journal of Intelligent Systems and Applications, 16(2), 45–58.

Bartsch, H., et al. (2024). Self-consistency of large language models under ambiguity. arXiv. https://arxiv.org/abs/2403.11234

Baur, D., et al. (2025). Development and evaluation of a retrieval-augmented generation–enhanced chatbot for patient-centered information in orthopedics and traumatology. JMIR AI, 4. https://doi.org/10.2196/jmir.ai.2025

Caspi, A., Roberts, B. W., & Shiner, R. L. (2005). Personality development: Stability and change. Annual Review of Psychology, 56, 453–484.

Drobakha, A. (2023, December 23). TestPersona. Test: “Persona. What is my personality type” (TXu 2-408-293) [U.S. copyright registration]. Opulentia SC LLC.

Drobakha, A. (2024, May 1). PersonaMatrix Project. Stage 1: “Persona Methodology” (TXu 2-427-005) [U.S. copyright registration]. Opulentia SC LLC.

Drobakha, A. (2025, March 25). PersonaMatrix Game Framework and AI Architecture (TXu 2-441-123) [U.S. copyright registration]. Opulentia SC LLC.

Drobakha, A., Kalitkin, M., Klymenko, K., Nayda, R., Lahuta, L., & Kostenko, O. (2026a). Psychoactive triggers as a stimulus battery for measuring large language models (LLMs): A bridge between psychometrics, clinical psychology, and LLM engineering. Metaverse Science, Society and Law, 2(1). https://doi.org/10.69635/mssl.2026.2.1.31

Drobakha, A., Lahuta, L., Aponchuk, A., Kalitkin, M., & Nayda, R. (2026b). Psychological testing as an instrument of differentiated support in education, healthcare settings, and crisis life transitions: Typological, trait-based, and psychodynamic approaches. Metaverse Science, Society and Law, 2(2). https://doi.org/10.69635/mssl.2026.2.2.38

Drobakha, A., Raschupkina, D., & Lahuta, L. (2026c). LLM as the non-desiring Other: A psychoanalytic model of “Frozen Projection” and its operationalization within the PersonaMatrix framework. Metaverse Science, Society and Law, 2(2). https://doi.org/10.69635/mssl.2026.2.2.37

Gargari, O. K., & Habibi, G. (2024). Enhancing medical AI with retrieval-augmented generation: A mini narrative review. Medical AI Letters, 3(2), 114–120.

Hassanein, F. E. A., et al. (2025). Calibration of AI LLMs in dental education. Journal of Dental Education, 89(4), 412–425. https://doi.org/10.1002/jdd.13781

Hasvold, P. E., & Wootton, R. (2011). Use of telephone and SMS reminders to improve attendance at hospital appointments: A systematic review. Journal of Telemedicine and Telecare, 17(7), 358–364.

Huang, Y., et al. (2024). Calibrating long-form generations from large language models. arXiv. https://arxiv.org/abs/2402.05143

Jing, Z., et al. (2025). AI chatbots in pharmaceutical e-commerce. International Journal of Electronic Commerce in Healthcare, 11(1), 67–82.

Khneyzer, C., et al. (2024). AI-driven chatbots in CRM. Journal of Relationship Marketing, 23(2), 145–163.

Lamrhari, S., et al. (2024). A social CRM analytic framework. Information Systems and e-Business Management, 22(3), 311–329.

Li, Y., et al. (2025). Graph-based confidence calibration for large language models. IEEE Transactions on Knowledge and Data Engineering, 37, 512–526.

Libai, B., et al. (2024). Brave new world? On AI and the management of customer relationships. Journal of the Academy of Marketing Science, 52(1), 18–39.

McKinsey & Company. (2025). The future of wellness trends. McKinsey & Company.

McLean, S. M., Booth, A., Gee, M., Salway, S., Cobb, M., Bhanbhro, S., & Nancarrow, S. A. (2016). Appointment reminder systems are effective but not optimal: Results of a systematic review and evidence synthesis employing realist principles. Patient Preference and Adherence, 10, 479–499.

Medical Group Management Association. (2025). Patient no-shows in 2025: What’s changing and what to do about it. MGMA.

Neha, F., et al. (2024). Retrieval-augmented generation (RAG) in healthcare: A comprehensive review. Health Technology Assessment, 12(4), 210–228.

Ng Kok Wah, J. (2024). Hybrid AI-powered chatbots in healthcare. Healthcare Informatics Research, 30(2), 134–142.

Patel, L. (2024). AI enabled customer acquisition and retention. Journal of AI in Business and Marketing, 5(1), 22–37.

Pendyala, M. K., & Lakkamraju, V. (2023). Impact of artificial intelligence in customer journey. International Journal of Customer Relationship Management, 17(3), 88–103.

Street, R. L., Jr., Makoul, G., Arora, N. K., & Epstein, R. M. (2009). How does communication heal? Pathways linking clinician–patient communication to health outcomes. Patient Education and Counseling, 74(3), 295–301.

Subham, K. (2024). Integrating AI into CRM systems for enhanced customer retention. Journal of Marketing Automation, 8(2), 150–165.

Sun, G., & Zhou, Y. H. (2025). AI in healthcare: Digital communication. Journal of Medical Systems and Communication, 19(1), 40–55.

Xiong, G., et al. (2024). Benchmarking retrieval-augmented generation for medicine. Medical Image Analysis and Augmented Reality, 14(3), 175–190.