ABSTRACT: Speech is one of the most natural and efficient forms of communication among humans, and it is becoming increasingly integral to human-computer interaction in modern applications such as virtual assistants, smart devices, and automated customer service systems. Among the various technologies that leverage speech, speaker recognition stands out as a biometric method that identifies or verifies individuals based on the unique characteristics of their voice. This approach has gained significant traction due to its non-intrusive nature, ease of integration, and applicability across various domains such as security, forensics, and personalised user experiences. In this paper, we introduce a novel speaker recognition system specifically designed to identify speakers based on utterances in the Marathi language, a linguistically rich and widely spoken regional language in India. The system utilises Mel-Frequency Cepstral Coefficients (MFCCS) to extract distinguishing vocal features that closely mimic the human auditory perception. MFCCS are particularly effective in capturing speech’s phonetic and acoustic properties, making them a preferred choice in speech and speaker recognition tasks. Vector quantisation (VQ) is applied to optimise the feature set and reduce computational complexity. VQ compresses the high-dimensional MFCC feature vectors into representative clusters without significantly compromising accuracy, thereby enhancing system efficiency. These features are then processed using hidden Markov Models (HMMS), which are adept at modelling temporal sequences and dynamic variations in speech. HMMS offer a statistical framework that effectively captures the sequential nature of speech patterns, leading to more reliable speaker modelling and recognition.

ABSTRACT: Modern energy platforms orchestrate dozens of external APIs-grid market interfaces, fleet telematics, billing systems, weather services, regulatory reporting endpoints-into complex workflows that power demand response, fleet charging, energy trading, and customer operations. As these platforms scale to manage thousands of EVs across multiple grid markets, the orchestration layer itself becomes a critical bottleneck: manual workflow configuration is slow, rule-based error handling misses cascading failures, and static API routing wastes both latency budget and cost. This paper presents an AI-powered API orchestration engine that replaces rule-based workflow management with eight specialized ML models: an intent classifier that routes incoming requests, a graph neural network workflow planner that generates optimal API call sequences, an anomaly detector that predicts API failures 15 minutes before they occur, an RL-based SLA optimizer that dynamically tunes timeouts and retries, and four additional models for cost allocation, schema mediation, failure prediction, and load balancing. Through a 16-month production deployment managing 42 external API integrations and 85 internal microservices for a fleet of 10,200 EVs across four grid ISO markets, the platform achieves a 98.7% workflow success rate (up from 72.4% with manual orchestration), reduces average workflow execution time from 42 minutes to 2.4 minutes, auto-resolves 94.2% of API errors with a mean time to recovery of 12 seconds, and reduces per-workflow cost from $18.50 to $1.15-a 93.8% reduction. The platform manages $45.6M in annual energy transaction revenue and achieves break-even ROI in 7 months. Five energy vertical case studies validate cross-domain effectiveness.

Privacy-enhancing technologies (PETs) play a crucial role in the digital age, particularly for organizations seeking to comply with the stringent privacy regulations in other regions of the world. The article provides insight into the role of PETs, including encryption, anonymization, and blockchain, in managing and protecting personal data, as well as ensuring compliance with regulations such as the GDPR and CCPA. Since it is carried out using qualitative research, as evaluated in the article, the success and implementation of PETs in different industries can be analyzed through case studies, surveys, and interviews. The results indicate that although PETs are complementary in promoting privacy compliance to minimize data breaches, organizations are more likely to face difficulties when implementing them because they are very costly, technically challenging, and require changes in accordance with regulations. Based on this understanding, this article presents practical suggestions that business entities can apply to embrace PETs and enhance their data security efforts to mitigate the evolving regulatory environment. The purpose is to assist organizations in overcoming these obstacles and perfecting their strategy on compliance and data protection.

This defect detection in printed circuit boards (PCBs) is crucial to ensure reliability and functionality of the equipment used in all Industries. Based on the analysis, this paper develops a new PCB defect detector composed of Swin Transformer and synthetic data generation based on ControlNet. Lack of availability of data and expensive labeling process makes the traditional method of defect identification unproductive. To overcome these challenges, this study uses ControlNet to create relevant synthetic defect samples that enrich the dataset and enhance generalization. Experimental findings show high increase in performance from 87.5% to 93.4% in accuracy; 85.2% to 91.8% in precision and 83.6% to 91.2% in F1-score by integrating synthetic data. The proposed approach had better performance than baseline models, including CNN and YOLOv5, especially in detecting complex and rare defects. The findings of this research also serve the purpose of enhancing the reliability and efficiency of PCB inspection to reveal the effectiveness of using more sophisticated methods of data augmentation and transformer-based models in industrial problems.

ABSTRACT: Cities increasingly rely on citizen reporting (e.g., 311 service requests), IoT sensors, and open data portals to guide infrastructure maintenance and sustainability actions. Yet most deployments optimize service metrics in isolation—focusing narrowly on response time or backlog—without jointly considering equity across neighborhoods and sustainability outcomes such as emissions, resilience, and safety. This paper presents Civic‑Sustain AI, an equity‑constrained, multi‑objective framework that integrates (i) citizen engagement modeling, (ii) sustainability impact modeling, (iii) fairness constraints with subgroup auditing, and (iv) privacy‑conscious governance aligned with NIST AI RMF and the NIST Privacy Framework. We define two composite metrics—Citizen Engagement Score (CES) and Sustainability Impact Score (SIS)—and pose portfolio selection of municipal actions as a constrained multi‑objective Bayesian optimization problem using q‑Expected Hypervolume Improvement (qEHVI). We articulate a reproducible evaluation plan on Open311‑compatible data (e.g., NYC 311), EPA EJScreen indicators, climate exposures, and DOT asset records, and provide an illustrative synthetic study to characterize trade‑offs on the CES–SIS Pareto frontier. Civic‑Sustain AI aligns outputs with ISO 37120/37122 [4,5] indicator families so that optimization results map to reporting standards used by cities. We discuss risk, privacy, and ethical guardrails; threats to validity; and pathways to deployment via digital‑twin dashboards. A brief, non‑central reference to the DICE framework is included to motivate multi‑source data‑integration concepts while maintaining a primarily civic focus

ABSTRACT: The evolution of artificial intelligence has reached an inflection point where raw computational power and isolated algorithmic advances are insufficient to address the complexity of real-world problems. This paper introduces and explores the synergistic paradigm of Machine Intelligence (MI)—encompassing not just machine learning, but also reasoning, adaptation, and autonomous decision-making—powered by Smart Data Technologies (SDT)—systems that actively curate, enhance, and govern data throughout its lifecycle. We posit that the next leap in capability will arise from their deep integration, creating systems where intelligence is embedded within a selfoptimizing data ecosystem. A comprehensive literature survey traces the lineage from data mining and business intelligence to contemporary deep learning and data-centric AI, establishing smart data as the essential substrate for robust machine intelligence. We propose the Intelligent Data Fabric (IDF), a novel conceptual architecture that unifies autonomous data pipelines, selfdescribing metadata (data intelligence), and adaptive machine intelligence models into a single, cognitive data plane. To validate this framework, a mixed-methodology approach was employed. First, a controlled simulation of a smart city traffic management system was developed, comparing a traditional data warehouse + ML model approach against an IDF-powered system. Second, a real-world case study of a financial fraud detection platform transitioning to an MI-SDT architecture was analyzed. The simulation results demonstrated that the IDF system achieved a 42% faster anomaly detection response time and a 35% reduction in false positives by dynamically retraining models on freshly curated, context-rich data streams. The financial case study revealed a 60% decrease in time-to-insight for new fraud patterns and a 50% reduction in data preparation overhead. Crucially, the study identifies metacognition—the system's ability to monitor its own data quality, model performance, and knowledge gaps—as the critical emergent property of this integration. The analysis concludes that the future of enterprise and societal-scale AI lies in moving from 'big data' to 'intelligent data,' where data is not passively stored but actively participates in the learning and reasoning loop. Key research frontiers include neuro-symbolic integration within data fabrics, federated intelligence across distributed smart data silos, and the development of ethical frameworks for autonomous data curation. The MI-SDT convergence represents a foundational shift towards truly autonomous, resilient, and context-aware intelligent systems.

ABSTRACT: Data mining has made supply chain management (SCM) a new research area where business organisations can gain a competitive advantage, and play a key role in the organisations management decision. The fundamental nature of Data mining usage in business applications is that it is useful in identifying relevant information and knowledge which can be used for better decision making in business area. This paper studies the use of data mining, as a tool, in understanding supply chain management (SCM). This also happens to generate large amount of data in its operations. Organizations can make use of this data for planning merchandise assortments, pricing the merchandise, vendor performance analysis, inventory control, product movement & supply chain and demand forecasting

The rapid adoption of cloud computing has introduced new security concerns, particularly regarding data privacy and protection from sophisticated cyberattacks. Intrusion Detection Systems (IDSs) based on Machine Learning (ML) offer improved detection of malicious behaviour in network traffic. However, existing IDS solutions struggle with large-dimensional data and imbalanced datasets, resulting in higher false positive rates (FPR) and reduced detection rates (DR). In this paper, we propose an improved Cloud IDS leveraging a hybrid features selection approach using Information Gain (IG), Chi-square (CS), and Particle Swarm Optimization (PSO). To address the challenge of imbalanced datasets, the Synthetic Minority Over-sampling Technique (SMOTE) is employed. The Random Forest (RF) classifier is used to detect and classify attacks. The system is evaluated on the UNSW-NB15 and Kyoto datasets, achieving accuracies of over 98% and 99%, respectively, outperforming other methods in terms of detection accuracy.

The proliferation of Internet of Things (IoT) devices and networks has led to an exponential increase in data generation at the network edge. Processing this data in real-time to enable rapid decision making presents significant challenges for traditional cloud-centric architectures. Edge AI, which involves deploying artificial intelligence algorithms directly on edge devices and gateways, has emerged as a promising solution to enable lowlatency analytics and decision making in IoT networks. This paper presents a comprehensive review and analysis of Edge AI techniques for real-time decision making in IoT environments. We examine the key components, algorithms, and architectures for Edge AI systems, as well as the challenges and opportunities in this rapidly evolving field. Through extensive experiments and case studies, we demonstrate how Edge AI can significantly improve response times, reduce bandwidth usage, enhance privacy, and enable new IoT applications across multiple domains including smart cities, industrial IoT, and autonomous systems. Our results show that Edge AI can reduce decision-making latency by up to 90% compared to cloud-only approaches while maintaining comparable accuracy. We conclude by discussing future research directions and the potential impact of Edge AI on the continued growth and evolution of IoT networks.

SAP S/4HANA Cloud represents a transformative evolution in enterprise resource planning systems, offering comprehensive cloud-native architecture that leverages advanced containerization, microservices, and modern DevOps methodologies. This article examines the platform's capabilities across multiple dimensions, including system architecture, performance optimization, multi-cloud deployment strategies, and business value realization. The article explores how the platform's in-memory computing capabilities, intelligent automation features, and advanced analytics functionalities deliver enhanced operational efficiency and strategic advantages. Through detailed examination of implementation considerations, security frameworks, and integration capabilities, this article demonstrates how SAP S/4HANA Cloud enables organizations to achieve digital transformation while maintaining robust security and compliance standards.