Smart Public Service Delivery through Telepresence and AI-powered Dashboards

Thesis Abstract

The rapid digitization of public administration and the growing demand for citizen-centric services necessitate innovative approaches to service delivery that combine telepresence technologies with intelligent data visualization. This study investigates how Telepresence-based service interfaces integrated with AI-powered dashboards influence the efficiency, transparency, and user satisfaction of public service delivery in metropolitan government offices. The central aim is to determine whether telepresence-enabled interactions coupled with real-time analytics dashboards improve service accessibility, reduce processing times, and enhance perceived legitimacy and accountability. Specific objectives include (1) evaluating the impact of telepresence on wait times and service completion rates; (2) assessing citizens’ perceived trust, transparency, and satisfaction with AI-powered dashboards; (3) identifying the organizational capabilities and governance conditions necessary to sustain telepresence-enabled services; (4) examining any differential effects across demographic groups and service types; and (5) developing a parsimonious conceptual model linking telepresence usage, dashboard analytics, and service outcomes. A mixed-methods research design is employed, combining a quasi-experimental field experiment with a concurrent explanatory sequential component. The population comprises three municipal departments piloting telepresence kiosks and dashboards in a major city. A sample of 12 service lines (e.g., licensing, permit issuance, and welfare inquiries) is selected, with 6 departments implementing the telepresence-AI dashboard intervention and 6 serving as matched controls. The study targets a total of 1,200 citizen-service encounters over three months, with 600 encounters in intervention sites and 600 in control sites. Data collection instruments include (i) operational metrics from enterprise service systems (average handling time, queue length, resolution rate, and follow-up requests); (ii) citizen surveys measuring perceived ease of use, trust, transparency, satisfaction, and perceived fairness; (iii) staff surveys assessing workload, perceived usefulness, and training adequacy; (iv) semi-structured interviews with policymakers, IT managers, and frontline officers; and (v) system logs from telepresence devices and dashboards capturing interaction patterns and anomaly events. Validity and reliability are ensured through pilot testing, Cronbach’s alpha checks for multi-item scales, and triangulation across quantitative and qualitative data. Quantitative analysis employs regression-based techniques to quantify the impact of the intervention on service efficiency and citizen satisfaction, controlling for client demographics, service type, and time-of-day effects. A difference-in-differences (DiD) approach is used to isolate treatment effects, supported by propensity score matching to address selection bias. Multivariate ANOVA tests explore interaction effects between service type and intervention, while structural equation modeling (SEM) tests the hypothesized pathways linking telepresence experiences, dashboard trust, and perceived service legitimacy. Qualitative data are analyzed through thematic analysis guided by the Technology Acceptance Model and legitimacy theory, with coding validated by intercoder reliability and member checks. A convergent parallel design ensures integration of quantitative and qualitative findings for a robust interpretation. Expected findings indicate that telepresence coupled with AI dashboards reduces average service times by 18–25%, lowers customer wait times by 22–30%, and increases first-contact resolution by 12–15%, relative to controls. Citizen trust, perceived transparency, and overall satisfaction are anticipated to rise significantly, mediated by perceived system usefulness and ease of use. Staff reports are expected to reveal improved workload balance and enhanced decision support, provided adequate training and governance structures are in place. The study is expected to reveal differential benefits across service categories, with more complex regulatory services showing larger efficiency gains but requiring greater human-in-the-loop oversight. The contribution to knowledge includes (a) empirical evidence on the effectiveness of integrated telepresence and AI-powered dashboards in public administration, (b) a validated conceptual model linking technological interfaces to service outcomes and legitimacy, and (c) actionable governance recommendations for scalable deployment, data governance, privacy, and ethical considerations. The study concludes that carefully designed telepresence interfaces, when paired with transparent, real-time dashboards and strong organizational capabilities, can substantially improve public service delivery while enhancing citizen trust. Policy implications emphasize investing in interoperable IT architectures, standard operating procedures for remote interactions, continuous training, and robust monitoring frameworks to sustain improvements beyond pilot implementations. Recommendations for future research include longitudinal studies across diverse municipalities, examination of equity implications for digitally underserved populations, and exploration of advanced analytics such as causal inference methods to further elucidate mechanism pathways.