Category Archives: Uncategorized

Authors: Mrs. P. Valarmathi, S.Abilesh, K.Karthick, T.Manoj

Abstract: In the current digital health ecosystem, users often rely on multiple fragmented applications for food tracking, nutrition analysis, and fitness planning, leading to poor user experience, limited personalization, and reduced adherence. This project proposes an AI-Powered Smart Diet and Workout Assistant, a unified web-based platform that integrates diet planning, calorie tracking, recipe generation, and workout recommendations into a single, personalized system. Users securely register, set health goals, and receive tailored plans based on their profiles, with AI-driven food recognition from images or text inputs, nutritional estimation, and deep learning models for diverse cuisines. Built with HTML/CSS/JS frontend, Node.js backend, MongoDB, and TensorFlow, it features progress dashboards, quizzes, and motivational tools. The system enhances engagement, consistency, and long-term health outcomes by minimizing app fragmentation and delivering intelligent, interactive fitness support.

DOI: https://doi.org/10.5281/zenodo.19216771

Published by: vikaspatanker

Authors: Thenmozhi P, Abarna M, Mahalakshmi D, Malini S

Abstract: Hybrid and nearshore outsourcing models are widely used to balance cost efficiency, talent availability, and operational flexibility, but they face challenges such as time-zone misalignment, uneven workload distribution, and limited performance monitoring. Traditional project management tools rely on static coordination and lack intelligent decision-making. This work proposes a smart platform based on an agentic AI-driven multi-agent architecture to manage distributed teams. The system decomposes project goals into tasks and assigns them using expertise, time-zone compatibility, and historical data. Specialized AI agents handle scheduling, performance prediction, and risk assessment. Built on an event-driven architecture, the platform enables real-time synchronization and continuous learning. Results show improved task allocation, early risk detection, and enhanced productivity compared to traditional approaches.

DOI: https://doi.org/10.5281/zenodo.19216630

 

Published by: vikaspatanker

Authors: Haseja Monika, Rathod Nidhi, Prof. Harkishan Gohil

Abstract: Software maintenance is one of the most cost-intensive phases in the software development lifecycle. A prevalent and paradoxical phenomenon — wherein the act of fixing a defect inadvertently introduces one or more new defects — significantly undermines software quality and reliability. These newly introduced defects, commonly termed regression bugs or fix-inducing changes, account for a substantial portion of post-release failures. This paper presents a comprehensive review of the causes, patterns, and mitigation strategies associated with bug-fix- induced regressions. We examine the theoretical foundations of software coupling and co-change dependencies, analyze empirical studies across open-source and industrial codebases, and survey state-of-the-art techniques including regression test selection, change impact analysis, automated patch validation, and AI-assisted code review. Our review identifies that insufficient test coverage, poor change impact analysis, high code coupling, and developer cognitive overload are the primary contributors to regression introduction. We further discuss the role of technical debt and architectural erosion in amplifying this phenomenon.

DOI: https://doi.org/10.5281/zenodo.19215822

 

Published by: vikaspatanker

Authors: Shekar Vollem

Abstract: Modern distributed computing environments support critical digital services but frequently encounter operational instability caused by complex interdependencies, infrastructure failures, and delayed incident response. These challenges highlight the need for intelligent infrastructure systems capable of identifying anomalies early and initiating automated corrective actions without human intervention. This study investigates the development of an autonomous self healing infrastructure framework that integrates predictive monitoring with intelligent automation to strengthen reliability, resilience, and operational continuity across distributed computing platforms. The research addresses the problem of reactive infrastructure management by proposing a proactive model that continuously analyzes operational telemetry, predicts potential system failures, and triggers automated remediation workflows. A mixed methodological approach is adopted, combining quantitative analysis of system performance metrics with qualitative evaluation of automation effectiveness in simulated distributed infrastructure environments. Predictive models analyze infrastructure signals such as resource utilization patterns, system logs, and service latency to detect early indicators of degradation, while automation components coordinate corrective responses including resource reconfiguration, service restart, and workload redistribution. Experimental observations indicate that the proposed framework significantly reduces incident response time, improves system availability, and enhances infrastructure stability during abnormal operating conditions. The findings demonstrate the strategic value of predictive automation in enabling autonomous infrastructure operations and minimizing manual intervention. This research contributes to the advancement of resilient infrastructure engineering by providing a scalable framework that supports proactive infrastructure management and strengthens reliability across complex distributed computing ecosystems.

DOI: https://doi.org/10.5281/zenodo.19208689

Published by: vikaspatanker

Authors: Shekar Vollem

Abstract: Enterprise cloud applications are increasingly required to support rapid software delivery, continuous updates, and highly reliable deployment cycles in order to meet the growing demands of digital transformation, global scalability, and user expectations for uninterrupted services. Continuous Integration and Continuous Delivery (CI/CD) pipelines have emerged as critical infrastructure components that enable automated building, testing, and deployment of applications in modern DevOps environments. These pipelines integrate development, testing, and operational workflows, allowing software changes to be validated and deployed in a consistent and repeatable manner. However, large-scale enterprise systems face significant challenges in optimizing CI/CD pipelines due to complex application architectures, distributed development teams, microservice dependencies, heterogeneous cloud infrastructures, and stringent compliance or security requirements. Inefficient pipelines can introduce bottlenecks in build processes, increase testing overhead, and slow down deployment cycles, thereby affecting overall software delivery performance. This paper explores strategies for optimizing CI/CD pipelines in enterprise cloud environments, focusing on automation frameworks, pipeline orchestration mechanisms, intelligent test management, infrastructure-as-code practices, and scalable deployment models that support cloud-native architectures. By analyzing existing research studies, DevOps methodologies, and industry practices, the study highlights architectural patterns, deployment pipeline designs, and continuous engineering principles that enhance the efficiency, scalability, and reliability of software delivery systems. The findings demonstrate that optimized CI/CD pipelines significantly improve release velocity, enable faster feedback loops for developers, reduce operational risks associated with manual deployments, and support scalable cloud-native application development while maintaining high standards of software quality and system stability.

DOI: https://doi.org/10.5281/zenodo.19208630

Published by: vikaspatanker

Authors: Dr. Saroj Agarwal, Sumit Sharma, Tanmay Kumawat, Vikas Bansal

Abstract: Traditional attendance monitoring systems rely heavily on manual processes or contact-based biometric solutions, which often lead to inefficiencies, proxy attendance, and lack of real-time behavioural insights [7]. Recent advancements in computer vision [5] have introduced facial recognition-based attendance systems; however, most existing solutions focus only on identity verification and fail to analyze participant engagement or emotional response during sessions [6]. This paper presents a comprehensive review and analysis of an integrated Facial Attendance and Sentiment Tracking System (FASTER), which combines real-time face detection [1], facial recognition using LBPH [2] and SVM classifiers [3], and expression-based sentiment monitoring [6] within a lightweight client-server architecture. Unlike previous systems that utilize either attendance automation or emotion detection independently, the proposed approach integrates both functionalities using OpenCV-based face detection [8], machine learning classifiers, and real-time data logging mechanisms. The system emphasizes low computational overhead, offline ca- pability, and user-friendly GUI-based interaction, making it suit- able for educational and organizational environments. Through comparative analysis with existing research, this study identifies key limitations in prior work and highlights the novelty of a unified attendance and sentiment-aware monitoring framework.

Published by: vikaspatanker

Authors: Deepa M P

Abstract: In the digital era, data is considered a valuable asset for organizations and industries. Database Management Systems (DBMS) provide a systematic way to store, manage, and retrieve data efficiently. With the rapid growth of technology, DBMS has become essential in integrating operations across various sectors. From banking transactions to healthcare records and e-commerce platforms, databases play a crucial role in ensuring seamless functionality and decision-making. Database Management Systems (DBMS) have become a fundamental component in modern digital infrastructure, enabling efficient storage, retrieval, and management of data across diverse sectors. This paper explores the role of DBMS as a core technology integrating multiple domains such as banking, healthcare, education, e-commerce, and government systems. It highlights how databases ensure data consistency, security, and scalability while supporting real-time applications. The study also examines emerging trends such as cloud databases, AI integration, and distributed systems. The findings demonstrate that DBMS acts as a unifying backbone, driving digital transformation and improving operational efficiency across sectors.

DOI: https://doi.org/10.5281/zenodo.19207441

Published by: vikaspatanker

Authors: Aditi Gaur

Abstract: Workplace harassment and gender inequality continue to be persistent challenges in urban institutions despite increasing female participation in the workforce and the presence of legal safeguards. This paper examines the nature, forms, and impact of workplace harassment on women employees in urban public institutions. It also explores how structural inequalities, patriarchal norms, and organizational culture contribute to gender-based discrimination. Drawing on sociological theories and existing literature, the study highlights the gap between policy and practice, particularly in the implementation of laws such as the POSH Act. The paper concludes that while urban institutions offer better employment opportunities, they also reproduce gender inequalities through subtle and overt mechanisms. Policy recommendations are provided to promote safe and inclusive workplaces.

DOI: https://doi.org/10.5281/zenodo.19205342

 

Published by: vikaspatanker

Authors: Aryan Nandgaonkar, Prathmesh Kore, Mayur Godse, Om Dhamale, Shital Kawale

 

Abstract: EDUFLOW is an advanced, AI-powered educational management system designed to enhance the learning and teaching experience by integrating modern technologies with intelligent automation. The primary objective of the system is to simplify academic processes such as content creation, assessment generation, timetable management, and resource organization for both students and teachers. Traditional educational systems often face challenges such as time-consuming content preparation, lack of personalized learning support, and inefficient resource management. EDUFLOW addresses these issues by providing a centralized platform that leverages artificial intelligence to automate and optimize educational tasks. The system enables students to generate study materials, practice quizzes, and personalized timetables, helping them improve their learning efficiency and time management. At the same time, teachers can create quizzes, exams, and teaching schedules with minimal effort, reducing their workload and allowing them to focus more on effective teaching. One of the key features of EDUFLOW is its integration with AI models, which generate high-quality educational content such as multiple-choice questions, study notes, flashcards, and summaries based on user input. This significantly reduces manual effort and ensures the availability of diverse and up-to-date learning resources.

DOI: https://doi.org/10.5281/zenodo.19205066

 

Published by: vikaspatanker

Authors: Imrana. Z, Sanjay. S, Dr. K. Brindha

Abstract: Healthcare monitoring systems are evolving rapidly with the integration of artificial intelligence, wearable sensors, and cloud-based data analytics. Traditional healthcare monitoring approaches rely on periodic medical examinations which may fail to detect early health risks. This research proposes an AI-assisted predictive health monitoring framework capable of analysing physiological data collected from wearable devices. The system processes health indicators such as heart rate, sleep patterns, and physical activity to identify abnormal trends and provide early alerts. Machine learning algorithms are employed to analyse patterns and support preventive healthcare monitoring. Experimental evaluation indicates that predictive analytics improves early health risk detection compared to conventional monitoring approaches. The proposed system highlights the importance of integrating intelligent analytics with digital healthcare systems.

DOI: https://doi.org/10.5281/zenodo.19203957

Published by: vikaspatanker