Browsing by Author "Vaz, Sonia"

Now showing 1 - 10 of 10
  • Thumbnail Image
    Item
    A Generative AI-Driven Cloud Framework for Predictive Optimization in Smart Manufacturing Ecosystems
    (Innovations in Machine, Engineering, and Digital Conference (IMED) (pp. 1-6), 2026) Vaz, Sonia
    The ecosystems of smart manufacturing produce huge quantities of heterogeneous data through sensors, machines, and manufacturing processes that provide the prospects of predictive optimization and present the challenges associated with data scarcity and uncertainty. Conventional predictive models tend to use very few historical failure data and deterministic forecasts and therefore become less effective in dynamic industrial processes. The paper suggests a cloud-based architecture based on the generative AI model that combines generative modelling, predictive analytics, and scenario-based optimization to improve decision-making in smart manufacturing. Generative AI is used to simulate more real operational conditions in the future, such as infrequent failure cases, whereas deep learning models predict the health of the machine, its production rate, and quality. These projections are converted into prescriptive operations with the help of a cloud-based stochastic optimization engine that considers the limits of operations and uncertainty. Empirical test on an industrial testbed shows that the framework proposed is able to drastically decrease downtime and unplanned failures and increase throughput and energy efficiency when compared to rule-based systems and predictive only systems. The findings shape how effective is the integration of generative intelligence and cloud optimization of manufacturing performance enhancement due to its robust and scalable nature. The experimental results accelerate the productivity of nearly 22 percent, reduce energy consumption by 18 percent and result in a 28 percent decrease in unplanned downtime when compared to conventional optimization approaches. Furthermore, the results indicate that the framework is robust and adaptable to dynamic workload fluctuations and does not suffer from the impact of concept drift; thus, it is suitable for scalable and adaptive smart manufacturing environments.
  • Thumbnail Image
    Item
    A study of service quality in Indian public sector banks using modified SERVQUAL model
    (Cogent Business & Management, 2023) Vaz, Sonia
    Assessment of service quality has been widely utilized in the service sector, especially in the banking industry. The present study aims to understand the influence of service quality on customer loyalty in Indian public sector banks. The service quality is quantified with the help of a modified SERVQUAL model using dimensions Reliability, Assurance, Tangibles, Empathy, Responsiveness, Charges, and Convenience. Structural equation modelling (SEM) indicated that among all the dimensions, Assurance, Empathy, Responsiveness, and Tangible were found to have a significant relationship with service quality. The banks must focus on bringing in innovation in these parameters to maintain a high quality of service and achieve higher satisfaction, which subsequently develops customer trust towards the company. By bringing innovative changes to improve the service quality, the banks can also increase their competitive advantage and customer retention as service quality has a significant relationship with customer loyalty.
  • Thumbnail Image
    Item
    AI-Powered FinTech Analytics for Transactional Transparency and Fraud Mitigation in Industrial IoT Ecosystems
    (Innovations in Machine, Engineering, and Digital Conference (IMED) (pp. 1-7), 2026) Vaz, Sonia
    The intensive introduction of Industrial Internet of Things (IIoT) technologies has already increased the quantity and complexity of machine-to-machine financial transactions, which has become a significant challenge in terms of the clarity of the story and exclusion of threats of fraud. The idea proposed in this paper is AI-based FinTech analytics that feature both supervised and non-supervised machine learning systems when detecting fraud in industrial transactional environments. Advanced feature engineering and imbalance-sensitive model evaluation metrics are used to improve detection accuracy and reliability. Explainable artificial intelligence methods have been used to foster regulatory trust. The experimental findings show that there is a high rate of fraud detection, a low rate of false alarms, and applicability in real time, which validates the efficiency of the proposed approach in light of safe and transparent IIoT-based financial systems. The superior performance of XGBoost (97.6%) is attributed to the optimization of the gradient for fraud detection.
  • Thumbnail Image
    Item
    Artificial intelligence driven sentiment analysis and market intelligence for strategic business decision making
    (ES Food and Agroforestry, 2025) Vaz, Sonia
    There is an increasing threat to agricultural productivity and food security because of the problems posed by climate change, limited resources, and inefficient supply chain management. The purpose of this study is to introduce AI-based forecasting and supply optimization network (AIFSO-Net), which is an artificial intelligence-driven network for supply optimization and forecasting. Dynamic optimization techniques and hybrid deep learning Convolutional Neural Network-Bidirectional Long Short-Term Memory (CNN–BiLSTM) are the two components that constitute this system. By utilizing data from multiple sources, including IoT sensors, satellite imaging, and meteorological records, the system can make predictions regarding crop yields and improve resource allocation in real time. Through the implementation of Predictive Reinforcement Learning (PRL), AIFSO-Net can improve the efficiency of its supply chain by adjusting to changes in the environment and market. The experimental results reveal that AIFSO-Net outperforms existing models, achieving a 12.8% improvement in forecasting accuracy and a 15.6% increase in supply chain efficiency. This is compared to standard methods such as transformers and genetic algorithm-based models. AIFSO-Net can improve agricultural forecasting, reduce post-harvest waste, and increase food security systems worldwide. In conclusion, AIFSO-Net offers a solution for sustainable agriculture that is both scalable and adaptable in addition to providing vital information for decision-making in agricultural situations that are always changing. The novelty of this work lies in the unified integration of hybrid deep learning, multi-source data fusion, and adaptive optimization (Dynamic Multi-Objective Optimization Algorithm (DMOOA) + PRL) within a single end-to-end framework for both crop forecasting and supply chain management, an aspect not addressed in the existing literature. This study contributes to sustainable agriculture and responsible resource management through accurate forecasting and adaptive supply chain optimization in alignment with Sustainable Development Goals 2 and Sustainable Development Goals 12.
  • Thumbnail Image
    Item
    Federated and Explainable AI Models for Secure FinTech Transactions in Digital Manufacturing Supply Chains
    (Innovations in Machine, Engineering, and Digital Conference (IMED) (pp. 1-6), 2026) Vaz, Sonia
    Digital manufacturing supply chains are becoming increasingly dependent on inbuilt FinTech services to perform automated payments, invoicing, and settlements which presents sensitive financial and operational data to security and privacy threats. This article is an empirical paper concerning the application of Federated Learning (FL) and Explainable Artificial Intelligence (XAI) in securing FinTech transactions in decentralized manufacturing supply chains. The suggested framework will facilitate joint fraud and anomaly-related detection without exchanging raw data between supply-chain participants. Different privacy mechanisms such as client-level and secure aggregation are integrated to safeguard sensitive data and minimize the risks of inferences. Explainable AI methods are used such as SHAP, local surrogate models, to enable transparency and auditability as well as regulatory compliance. Experimental evidence has shown that federated models can attain almost centralized detection accuracy with much stronger privacy guarantees and explainability procedures can give insightful and interpretable information about model decisions. The paper identifies the trade-offs between accuracy, privacy, and computational overhead and concludes that federated and explainable AI provides a convenient, secure, and compliant solution to FinTech-enabled digital manufacturing ecosystems.
  • Thumbnail Image
    Item
    Impact analysis of e-learning on students of higher education institutions during COVID-19: A structural equation modelling approach
    (RECENT ADVANCES IN SCIENCES, ENGINEERING, INFORMATION TECHNOLOGY & MANAGEMENT 2782.1 (2023): 020198, 2023) Vaz, Sonia
    Classroom-based, face-to-face interactive teaching has been a conventional system for decades in higher education. Due to the spread of COVID-19, universities were forced to halt their educational programmes. As a result of integrating technology and education, e-learning has become a vital learning medium. As e-learning becomes progressively essential in education, there has been a significant increase in e-learning courses and programmes. E-learning systems play a critical role in today’s educational landscape and must be evaluated to ensure decisive delivery, pragmatic use, and a positive impact on learners. As a result of this extensive review of the existing literature and the development of a comprehensive model, different rates of success can be linked to different factors. The Technology Acceptance Model (TAM) and the User Satisfaction Model (USM) were both used to support our findings. PLS-SEM (Partial Least Squares–Structural Equation Modelling) was used to analyze data from 352 students who were participating in an e-learning course. Using this model, the study describes how learners’ self-regulation mechanisms and attitudes, variations in temperament, and extrinsic considerations such as technical assistance, development preparation, and usability of facilities influence the perceived ease of use and perceived value of electronic learning programmes
  • Thumbnail Image
    Item
    Impact analysis of e-learning on students of higher education institutions during COVID-19: A structural equation modelling approach
    (Journal of Computers, Mechanical and Management, 2022) Vaz, Sonia
    Staff members use tried-and-true procedures when completing workplace visits, delivering services, and completing client tasks. However, the COVID-19 pandemic compelled employers to change the work styles of individual employees to ensure good communication, work-life balance, and flexibility for employees while maintaining optimal work productivity levels. In addition, the World Health Organization established social separation guidelines to combat COVID-19. Thus, the pandemic challenged the work culture and resulted in employees being quarantined in their homes. As a result of this transformation, employees were encouraged to use digital tools to facilitate work-from-home opportunities. The current study analyzes employees' psychological and productive effects of work-from home culture. It also looks for coworker bonding threatened by this transformation and suggests a way to keep it intact. Through a thorough literature review, the authors developed a comprehensive model to assess the pandemic's impact on employees' lifestyles. The conceptual model was empirically tested by applying the model to data collected from 233 employees from various backgrounds. The model result was validated using Partial Least Squares Methods-Structural Equation Modeling. The inferences highlight the factors influencing employee morale and work culture and the parameters closely related to employee functioning in the organization that should not be affected.
  • Thumbnail Image
    Item
    Leveraging ChatGPT for Empowering MSMEs: A Paradigm Shift in Problem Solving
    (Engineering Proceedings, 2024) Vaz, Sonia
    This paper delves into the potential of harnessing ChatGPT, an AI-driven language model, to empower micro, small, and medium enterprises (MSMEs) by revolutionising their approach to problem solving. The research aims to explore the integration of ChatGPT into MSME operations and evaluate its impact on enhancing their problem-solving efficiency. By scrutinising the literature and reviewing several case studies, a comprehensive framework emerges, detailing the utilisation of ChatGPT as a problem-solving tool for MSMEs. This involves training the model with industry-specific data and incorporating it into MSME communication channels, enabling intelligent responses to queries. The results highlight the substantial improvement in problem-solving capabilities, with the model’s real-time assistance diminishing response time, elevating accuracy, and furnishing tailored solutions to intricate challenges. However, limitations arise from the model’s reliance on existing data, potentially introducing biases. Significantly, this research offers practical implications for both MSMEs and policymakers. ChatGPT’s integration holds promise in terms of heightened efficiency, productivity, and competitiveness for MSMEs, counteracting resource constraints, and fostering growth. Policymakers can aid this transition by formulating ethical guidelines to ensure the equitable and transparent application of AI in the MSME sector. This study’s novelty lies in its focus on MSME empowerment through ChatGPT integration, bridging a research gap. Its value emanates from the actionable insights provided, offering guidance to MSMEs, policymakers, and practitioners keen on leveraging AI-driven solutions to amplify problem-solving capacities within the realm of MSMEs.
  • Thumbnail Image
    Item
    Redefining Workspaces: Young Entrepreneurs Thriving in the Metaverse’s Remote Realm
    (Engineering Proceedings, 2024) Vaz, Sonia
    This research paper explores the intersection of the Metaverse and remote working, specifically concerning young entrepreneurs. Its primary objective is to examine the opportunities and challenges presented by the Metaverse for this demographic engaged in remote work, providing actionable insights for both practitioners and policymakers. The methodology employed involves an extensive literature review that delves into the concept of the Metaverse, its evolution, and the implications it holds for remote working. This foundational exploration is supplemented by in-depth analyses of case studies and examples, offering real-life illustrations of how young entrepreneurs leverage the Metaverse for remote work. The findings of this investigation reveal a landscape ripe with potential for young entrepreneurs operating within the Metaverse. This study highlights the benefits of enhanced collaboration, expanded global market access, and the emergence of innovative augmented and virtual reality applications. However, these opportunities are accompanied by notable challenges, including issues related to technological infrastructure readiness, security concerns, and potential societal impacts. Acknowledging the evolving nature of the Metaverse concept and potential biases in sample selection are critical research limitations. Practically, this paper translates its findings into actionable recommendations for young entrepreneurs seeking to maximize their utilization of the Metaverse for remote work. It emphasizes the importance of skill acquisition, adaptability to the changing work environment, and the implementation of robust security measures. Furthermore, it advocates for policymakers to develop supportive regulations and policies that recognize and accommodate the intricacies of virtual contracts, data protection, and cross-border collaborations. Strengthening intellectual property laws and tailoring taxation policies for this digital domain are also crucial aspects. In essence, this research contributes significantly by synthesizing the existing literature, presenting real-world examples, and offering practical insights tailored to the unique space where the Metaverse and remote work intersect. Its value lies in bridging gaps in understanding, providing actionable guidance, and contributing to the evolving discourse on this emerging field.
  • Thumbnail Image
    Item
    Unlocking Brand Excellence: Harnessing AI Tools for Enhanced Customer Engagement and Innovation
    (Engineering Proceedings, 2024) Vaz, Sonia
    This research article delves into the integration of AI tools, particularly Chat GPT, within brand marketing strategies, aiming to uncover their practical applications and associated benefits and challenges. Real-world case studies, practical recommendations, and insights into AI-driven innovation collectively form a guide for brand managers aspiring to leverage these tools effectively. The research findings highlight Chat GPT’s transformative potential, showcasing successful integration into marketing strategies that enhance customer experiences, streamline interactions, and introduce innovative campaigns. Despite acknowledging the dynamic nature of AI technology and potential biases in data analysis, the article provides practical recommendations for brand managers, emphasizing ethical considerations and adapting to the evolving AI landscape. The research underscores the importance of responsible AI usage, transparency, and continuous adaptation to changing consumer behaviors for maintaining trust and ethical standards. This contribution to the existing literature combines real-world examples, practical insights, and a mixed-methods approach, offering a unique perspective on how AI, particularly Chat GPT, can reshape customer engagement, brand communication, and creativity in both academic and industrial contexts. The article provides a comprehensive examination of AI tools’ practical utility, bridging theory and application for a nuanced understanding in the field of brand marketing