A Secure Cloud–AI Collaborative Framework for Financially Intelligent and Autonomous Manufacturing Enterprises
Ieee
May 30, 2026
The paper is a proposal of secure cloud-AI cooperative model to empower financially intelligent and autonomous manufacturing businesses. The architecture combines Internet of Things data of industries, financial data of enterprises, and machine learning models on the cloud system. It involves single security platform that can be able to support scalable and reliable decision-making. Deterministic ML strategies are used to forecast financial risk, cost, and profitability. It can be further used by an independent decision engine to maximize manufacturing process. The experimental analysis by classification and regression indicators illustrates the presence of a better forecasting level, lower operational price, and effective mitigation of financial risks. The security analysis provides better data protection at minimal overheads of the system, which renders the framework to apply to the real-life intelligent manufacturing system. The current research proposes a Cloud–AI secure cooperating model as the basis for integrating security-aware cost optimization with a unified framework of adaptive resource orchestration and real-time governance using AI. The new solution is different from other current solutions (like FinOps) which treat costs or automation separately, but instead combine security, operation intelligence, and cost efficiency in one cloud-enabled manufacturing system.Show Less
Metadata
See publication
Tags: AI, IoT
Blockchain and AI Fusion for Financial Traceability and Cyber-Resilience in Industry 5.0 Environments
Ieee
May 30, 2026
The rapid process of digitalization of industrial finance in Industry 5.0 conditions demand increased transparency, intelligence, and cyber-resilience. In this paper, the authors suggest a hybrid Blockchain-Artificial Intelligence (AI) system of secure financial traceability and dynamic cyber resilience of industrial ecosystems. There is a high level of immutable and end-to-end traceability of financial transactions of blockchain technology, and machine learning models allow detecting fraud and predicting risk. Smart contracts can be used to improve compliance enforcement, and AI-based resilience modeling has a stronger role to play in the recovery of the system in the event of cyber threats. The suitability of the suggested methodology to the future industrial financial system is demonstrated by numerical data indicating that the confidence of traceability has significantly increased, the reliability of fraud detection, and the recovery time
See publication
Tags: AI, Blockchain, Cybersecurity
Multi-spectral Approach to Segment Remote Sensing Data
Springer Nature
April 30, 2026
One of the key challenges in precision agriculture is accurately distinguishing crops from the surrounding soil. Contemporary algorithms in precision farming rely on multi-spectral or hyper-spectral data and artificial intelligence networks for computing radiometric indices, supporting the operational management of agricultural systems. These transformations act as natural filters for multi-spectral and hyper-spectral imagery, reducing the complexity of data input while improving the network’s ability to classify information. This study suggests defining the radiometric index with the help of a directional mathematical filter in order to accurately segment crops from soil.
See publication
Tags: AI, Analytics, Emerging Technology
Deep Learning Framework for Detection of Neonatal Respiratory Abnormalities
Springer Nature
March 31, 2026
Respiratory diseases in neonates are among the leading causes of neonatal illness and death, particularly in developing countries. Prompt diagnosis and treatment of these conditions are essential. Thermal imaging emerges as a non-invasive and radiation-free diagnostic approach, utilizing temperature variations and thermal symmetry monitoring as tools in medical diagnostics. This study explores the detection of neonatal respiratory abnormalities using artificial intelligence applied to limited thermal imaging data. Convolutional Neural Network (CNN) models, while highly effective for classification tasks, typically require extensive and balanced datasets. However, obtaining sufficient neonatal thermal imaging data can be challenging due to the delicate nature of care in neonatal intensive care units. To address this limitation, the study incorporates a robust deep learning framework alongside various data augmentation techniques to enhance classification outcomes. Neonates with respiratory abnormalities were grouped into one category, while those with cardiovascular and abdominal issues were grouped into another. Results showed that data augmentation, which increased the dataset size by four times, improved classification accuracy from 84.5% to 90.9%. With deterministic feature extraction as well as the incorporation of supervised learning, this paper presents a hybrid method for data processing acquired through UAVs.
See publication
Tags: AI, Emerging Technology, Healthcare
Integrating Artificial Intelligence with Cybersecurity for Resilient Wireless Communication Against Advanced Threats
Ieee
September 30, 2025
In this paper, our focus is how AI could be integrated with cybersecurity solutions particularly to bolster the effectiveness of wireless communication systems in a progressive manner against sophisticated agile attacks. With the increasing need for secure and efficient wireless networks, AI-based solutions provide proactive threat detection, adaptive defence strategies, and performance optimization in the connected world. This study uses a range of experiments and analyses to show how AI can keep systems more stable, improve detection accuracy, and reduce false positives. AI plays a pivotal role between performance and strong cybersecurity, as shown in the spider and radar charts visualizing how the AI augments security services, regardless of the concern being high or low intensity. The results highlight the promise that AI holds to empower groundbreaking solutions for protecting next-generation wireless communication systems against emerging cyber threats
See publication
Tags: 5G, AI, Cybersecurity
A Security-First Approach to Federated Feature Engineering and Model Training in Cross-Cloud Ecosystems
IEEE
August 31, 2025
The increasing demand for raw data protection and organizational data lockdown conditions created an opportunity for federated learning (FL) to become an essential solution that enables collaborative training of machine learning models without exposing original data. The combination of different cloud environments in cross-cloud systems makes it hard to provide both data protection and consistent features. The proposed security-first federated framework incorporates secure multiparty feature engineering together with differential privacy mechanisms as well as homomorphic encryption into the FL lifecycle. The introduced approach implements zerotrust security protocols during inter-cloud communications and uses adaptive harmonization methods to maintain data usefulness. The proposed framework incorporates a privacy-protected federated training scheme which enables independent asynchronous model updates within distributed cloud systems to bolster protection against model corruption during training and inferential attacks. Analytical tests performed on authentic healthcare and financial system data revealed an 28% decrease in privacy breaches while raising training resilience by 34% and matching model precision levels relative to traditional FL systems. The research work forms an essential base for deploying trustworthy federated learning systems that enable secure and reliable operations between multiple cloud environments.
See publication
Tags: Cloud, Cybersecurity, Privacy
Federated Data Modeling for LLM Deployment in Secure Cloud-Native Architectures
Ieee
August 28, 2025
LLMs have brought new, amazing abilities for understanding language, generating it and making decisions. Yet, there are serious concerns about data privacy, the ability to scale LLMs and how different components of a cloud-native system interact. The paper outlines a new Federated Data Modelling (FDM) framework specifically for making use of LLMs in secure and efficient distributed cloud settings. The framework achieves decentralized training, prevents data being leaked and meets the requirements of data residency laws by using federated learning and dynamic schema harmonization with container orchestration. Moreover, the proposed FDM technique relies on zero-trust security, confidential computing and Kubernetes-native operations to provide isolation, watching and traceability among the various tenants. On typical benchmark datasets, the approach shown here performs better in terms of privacy, how quickly the model learns and how quickly it may be used in practice compared to centralized training. By using this study, AI service providers can ensure their LLM service is trustworthy and safe for IAP use in healthcare, finance and government
See publication
Tags: AI Infrastructure, AI Orchestration, Generative AI
Federated Learning Meets Data Engineering: Building Trustworthy AI with Decentralized Data Pipelines
IEEE
August 20, 2025
The increasing requirement for private machine learning models between organizations created rapid developments between federated learning and data engineering practices. The research investigates a new framework which combines trust mechanisms with decentralization methods and engineering discipline for the development of explainable and secure scalable AI systems. Data sovereignty and integrity together with verifiable model training are achieved through decentralized data pipelines which operate without exposing raw dataset information. The developed framework delivers reliable data processing methods which optimize federalized system operations and features trust-building features involving differential privacy implementation and secure aggregation capabilities alongside blockchain auditing capabilities. Our work includes the presentation of engineering methodologies for monitoring and orchestration as well as fault-tolerance techniques when deploying FL models across multiple heterogonous platforms. Results from experimental benchmark tests establish advancements in model accuracy performance and reductions in communication demands and improvements in trustworthiness compared to standard centralized and naïve federated learning models. The data indicates that industries which need strong privacy and transparency requirements should adopt engineering-centered federated learning systems as the basis for trustworthy artificial intelligence frameworks.
See publication
Tags: AI Governance, AI Orchestration, Blockchain
Predictive Analytics for Patient Readmission Using Machine Learning Models
Ieee
July 31, 2025
The use of machine learning models for predicting patient readmission risks in this paper is intended to improve healthcare decision. Model evaluation methods such as confusion matrix visualization, F 1 -score calculation, and SHAP (SHapley Additive Explanations) were used to understand the performance and explainability of the model. It shows the impact of changing the prediction thresholds on the prediction accuracy, while the SHAP summary plot illustrates the importance of predictors in predicting hospital readmission. Moreover, a comparative box plot indicating readmission risk across patient subgroups shows differences in demographics, which is important for targeted health care management. Machine learning models could provide new means of prediction to enhance patient outcomes, reduce hospital readmissions and guide precision
See publication
Tags: Health and Safety, Health and Wellness, Healthcare
AI-Driven Metadata Extraction and Classification using LLMs in Big Data Lakehouses
Ieee
July 31, 2025
Modern data lake houses encounter substantial challenges due to their rapid increase in heterogeneous and unstructured data when it comes to metadata administration and classification and semantic discovery. Such massive and diverse datasets exceed the limits of traditional schema-based methods as well as rigid rule-based systems. This research develops an AI-based method for intelligent classification and automated metadata extraction through Large Language Models. The system uses pre-trained LLMs' contextual understanding and transfer learning abilities to accurately extract metadata from documents logs and multimedia items which have semi-structured and unstructured formats. The framework incorporates a zero-shot learning method from LLMs which teams up with supervised fine-tuning for domainspecific tagging. The validated architecture shows better performance and adaptivity through achieving superior accuracy than traditional NLP-based metadata engines when applied to real-world lakehouse dataset storage. Experimental findings demonstrate that LLMs create substantial enhancements in metadata development as well as search functions and data protection metrics in modern big data systems. This proves LLMs can bring substantial changes to future big data platforms
See publication
Tags: AI Infrastructure, Big Data, Generative AI
IIOT Security Concerns: An Extensive Analysis of IIOT Attacks and Defences
IEEE
July 31, 2025
IIoT refers to the confluence of information technology and industrial processes, which might alter the industry. This convergence enables increased automation, data communication, and process optimization. However, this connection introduces additional security vulnerabilities and increases cyberattack risk. These consequences may affect the operational integrity and safety of many critical industrial processes. In this study, IIoT safety is thoroughly examined. The research covers all cyber dangers, from denial-of-service (DoS) and brute force to assaults. These models’ ability to identify and mitigate these risks’ detrimental effects is assessed using Random Forest, Decision Tree, and Gradient Boosting. Model efficacy is assessed using these models. We found that ecosystem protection for the Industrial Internet of Things (IIoT) requires a multi-layered, dynamic security strategy that considers human factors, technology solutions, and regulatory compliance.
See publication
Tags: Cybersecurity, IoT, Risk Management
AI-Enhanced Blockchain Networks for Climate Change Monitoring and Carbon Credit Verification
ACM
July 31, 2025
Climate change is the most pressing global problem, which warrants technological innovation in accurate monitoring and efficient market-based solutions. In this paper, we propose a framework to combine staking with artificial intelligence and blockchain to provide a transparent, secure, and efficient way of monitoring a variety of carbon credits related to carbon footprint. This uses machine learning algorithms to combine satellite imagery, IoT (wearable) data, and immutable blockchain ledgers to create tamper-proof environmental monitoring systems. It suggests brilliant contract architecture that can generate carbon credits through AI to validate the process, federated learning applications to track cross-border emission activity, and neural networks to validate carbon sequestration projects. Using these systems, we achieved orders of magnitude improvement in verification accuracy, transaction transparency, and market efficiency over traditional systems. By employing this integrated approach, some of the most pressing carbon market dilemmas, including narrowing carbon market data integrity issues, delays in verification, and deficits of trust among carbon market participants, can be resolved, and it is a strong foundation for climate action globally.
See publication
Tags: AI, Blockchain, Climate Change
Blockchain-Powered Smart Contracts for Automated Financial Transactions in Decentralized Networks
Ieee
July 30, 2025
This paper explores the application of blockchain enabled smart contracts to automate financial transactions across a decentralized network. We discuss performance measurements including transaction throughput, gas consumption, security, scalability, and the interoperability of cross-chain. We simulate traditional financial systems and blockchains and compare them in terms of transaction throughput, cost ratios and susceptibility of a smart contract to attacks. Furthermore, we devise the performance of a few types of blockchain platforms related to Ethereum and Binance Smart Chain under different network settings. Security issues, including integer overflow and re-entrancy, are mitigated using formal verification techniques. We also evaluate the impact of cross-chain interoperability in terms of bridges and oracles, with a specific emphasis on the latency and failure rate of the asset swaps. The results illustrate the capability of blockchain-based smart contracts to disrupt financial systems by providing a more efficient, secure, and scalable alternative
See publication
Tags: Blockchain, Cryptocurrency, FinTech
Natural Language Processing in Electronic Health Records for Automated Clinical Decision Support
Ieee
July 22, 2025
In this study, the aim is to research how NLP can be used to advance Clinical Decision Support Systems (CDS) in order to ensure better health outcomes for patients. This compares the performance of multiple NLP models across several metrics (including accuracy, F1-score, precision, and recall) such as BERT, BioBERT, GPT-4, LSTM, and CNN. The results show that the performance of the NLP-based systems surpassed the traditional rules, especially in NER and sentiment analysis tasks, that enable how fast we identify risks and diagnose. NLP also enhances clinical efficiency by greatly reducing diagnosis time. These results show the power of Natural language processing in relation to modernization of decision-making systems in healthcare domain
See publication
Tags: Analytics, Healthcare, HealthTech
Cybersecurity Challenges in AI-Driven Dynamic Spectrum Management for Next-Generation Wireless Systems
Ieee
July 01, 2025
This paper investigates intelligence-based spectrum management in one aspect in addition to the performance, and in another to security for next generation wireless systems. The research assesses how cybersecurity affects spectrum utilization, adversarial attack detection, system latency, and energy efficiency through simulated dynamic spectrum scenarios. The findings show that secure systems always outperform the insecure systems based on KPIs like spectrum efficiency, detection accuracy, and responsiveness of the system. In addition, secure setups will have much lower energy consumption because there is a direct trade-off between security and resources. This study demonstrates the necessity of integrating highly effective cybersecurity measures to protect against these growing threats, securing resilient and efficient space utilization in next generation wireless systems
See publication
Tags: AI, Cybersecurity, Telecom
Proactive Cybersecurity Frameworks using Machine Learning for Anomaly Detection in Wireless IoT Networks
IEEE
June 30, 2025
A Novel Proactive Cybersecurity Framework for Router-based Anomaly Detection in Wireless IoT Networks using Machine Learning Techniques It discusses the model complexity and energy-drain trade-offs, posing challenges to find a system that balances detection accuracy with minimal resource usage. The framework shows the adaptability of dynamic network conditions and resilience to adversarial attacks through extensive simulations. Federated learning also improves scalability and communication security as the number of IoT devices increases. Its analysis emphasizes devising strategic balance provisions between performance and efficiency to build scalable and secure IoT systems in constraint measure scenarios. This research is helpful in ushering smart, adaption, efficient and smarter way forward towards future IoT deployment.
See publication
Tags: AI Safety, Cybersecurity, IoT
Real-Time Cloud-Native Processing Techniques for Enhancing Federated Learning in Distributed Applications
IEEE
June 30, 2025
Federated Learning (FL) has become a new paradigm in the privacy-preserving domain for training machine learning models over the distributed clients without transferring raw data. But, the real FL offer challenges for dynamic and heterogeneous environments with the real-time data processing due to the reason of the environments being heterogeneous and dynamic. This paper offers a new design of a cloud-native architecture to overcome these barriers through scalable and efficient real-time data processing. The architecture combines microservices, container orchestration, and streaming frameworks for low-latency data and adaptive resource management. Implementing this framework in distributed cloud environments enables us to improve model convergence speed, system scalability, and performance in different federated learning scenarios. The experiments performed on benchmark datasets prove the approach to be better in terms of communication responsiveness and system responsiveness. This work provides a flexible and robust solution for deploying federated learning in constrained real-time, large-scale, resource-limited applications.
See publication
Tags: AI Infrastructure, AI Orchestration, Cloud
Hybrid AI Models for Privacy-Preserving Big Data Analytics in Distributed Environments
IEEE
June 30, 2025
many distributed environments that generate massive data such as edge devices and cloud platforms and IoT systems have led to the necessity of developing advanced big data analytics frameworks that combine performance with privacy safeguards. Machine learning systems constructed through traditional methods fail to achieve this dual requirement balance particularly within decentralized heterogeneous networks. This study introduces a new artificial intelligence system which unites federated learning and homomorphic encryption and swarm intelligence capabilities to secure distributed big data analytical computations. Through this proposed model sensitive data remains within its local space while the framework enables cooperative learning along with real-time operational procedures. Our framework achieves top accuracy on standard benchmark tests while ensuring low communication requirements and robust privacy protection. Research findings show hybrid artificial intelligence solutions can effectively resolve security and scalability problems which affect distributed systems analytics.
See publication
Tags: Analytics, Big Data, Privacy
Edge-Aware Federated Learning: A Scalable and Fault-Tolerant System Architecture
IEEE
June 30, 2025
Federated Learning (FL) enables decentralized machine learning by allowing edge devices to collaboratively train models without sharing raw data. This paradigm is particularly important in scenarios involving sensitive data and privacy concerns, such as healthcare, smart cities, and industrial IoT systems. However, the deployment of FL in edge environments is far from trivial due to critical challenges such as intermittent device connectivity, heterogeneous computational resources, non-independent and identically distributed (non-IID) data, and vulnerability to adversarial attacks. These constraints can severely impact the stability, scalability, and fault tolerance of FL systems, making robust solutions essential.To address these limitations, this paper proposes a scalable fault-tolerant framework for FL on the edge by combining robust aggregation, blockchain-enhanced trust, and federated continual learning (FCL). We evaluate the framework under simulated edge conditions and demonstrate improved reliability and model convergence.
See publication
Tags: AI Infrastructure, Blockchain, Smart Cities
Emerging Trends in Big Data Analytics and IoT for Autonomous Healthcare Technologies
IEEE
June 30, 2025
The convergence of Big Information Analytics (BDA) and the Internet of elements (IoT) which is defined by the use of Information-driven and Simplified Answers is changing patient care and the way healthcare is delivered. healthcare systems get better diligent outcomes Construct the world of customized discourse plans and further break decision-making away exploitation real-time information from connected devices. The present situation and potential Uses of these technologies in the future are examined within the purview of the research. the contributions that these technologies render to prophetic analytics far Watching and effective Productivity are apt particular condition. The advancement of proactive patient management has been fueled by the capacity of healthcare professionals to convert vast volumes of Information into pertinent Understandings via the use of advanced analytics and the Internet of elements. notwithstanding Problems care cybersecurity interoperability and information secrecy restrictions have work resolute inch rate to full see the call given away free healthcare systems.
See publication
Tags: Big Data, Healthcare, IoT
Deep Learning Based Edge Computing Model for IOT Data Analytics with Embedding Intelligence
IEEE
June 30, 2025
The rapid proliferation of linked devices, including wearables, sensors, mobile devices, and Internet of Things devices, has significantly increased network data traffic. Perhaps the source of this enormous amount of data is the Internet of Things. Data is sent to the cloud or another centralized system for machine learning (ML) via Internet of Things (IoT) devices. This method is still used even if network traffic and latency are rising. By moving processing duties to the network's edge, where data sources are located, edge computing may be able to solve these difficulties. The problem might be identified and fixed more quickly as a result. Machine learning is not a good fit for edge computing because of its limited processing power. The drawback of edge computing is substantial. Using edge nodes, this paper combines cloud and edge computing for IOT data analytics. Comparative analysis involves similarity-based processing. Deep learning could uncover novel characteristics. Perhaps the encoder is at the network edge, while the trained auto encoder's decoder is in the cloud. This study focused on the difficulties of detecting human activity from dwindling sensor data. Sliding windows in the deep auto encoder (AE) may reduce edge data by 80% during the preparation phase without compromising accuracy. With no impact on throughput, this is possible. Even without other methods, this is possible.
See publication
Tags: AI Infrastructure, Analytics, IoT
Secure LLM-Oriented Data Engineering Pipelines for Scalable AI Workflows in Multi-Cloud Environments
Ieee
June 10, 2025
Advances in LLMs have altered AI analytics, yet they are not yet easily brought into scalable, secure and efficient engineering systems—especially in settings with various cloud platforms. The paper offers a new design for LLM-driven data engineering processes that resolves major issues in collecting, processing and publishing data on the cloud. A proposed framework includes federated orchestration, encryption of all data and access control policies without trust, to achieve proper data privacy and adherence to rules, as well as ensure fast realtime processing of both unstructured and semi-structured data. In addition, we introduce a new scheduler and metadata manager to automatically optimize things like building prompts, training and inferencing using machine learning models. Extensive study using common benchmark sets on AWS, Azure and GCP demonstrates that there are 35% fewer pipeline problems, cost is reduced by 28% and the system performs more effectively and is more resilient. This research outlines how secure and scalable use of AI systems powered by LLMs can be spread across various multi-cloud platforms
See publication
Tags: AI Governance, AI Safety, Cloud
