Dr. Savya Sachi
Assistant Professor, Department of Computer Science
L. N. Mishra Institute of Economic Development and Social Change, Patna, Bihar
Dr. Savya Sachi is a dedicated academician with over 10 years of teaching experience across various esteemed institutions. Currently serving as an Assistant Professor in the Department of Computer Science at L. N. Mishra Institute of Economic Development and Social Change, Patna, he has made significant contributions to both teaching and research.
Dr. Savya Sachi has authored more than 30 research papers published in reputed journals indexed in SCI, SCOPUS, Web of Science, and UGC CARE. He has also written several academic books including Software Engineering, Computer Networks, and Programming for Problem Solving (PPS). In addition to his scholarly work, he holds multiple patents and copyrights.
His primary areas of expertise include Network and Web Security, Data Mining, Computer Networks, and Image Processing.
EDUCATION
BE(IT), M.TECH(CTA), Ph. D. CSE
RESEARCH, TEACHING, or OTHER INTERESTS
Computer Science, Computer Engineering, Artificial Intelligence, Computer Networks and Communications
Optimized Protected Data Communication in WSN Using Fusion Pretended Tempering with Lion Swarm Optimization and Improved Elliptic Curve Cryptography Saurabh Sahu, Anil Pratap Singh, Banothu Seva, Vandana Rathod, M. Shamila, Savya Sachi 2025 IEEE 5th International Conference on ICT in Business Industry and Government Ictbig 2025, 2025 A new Hybrid Simulated Annealing with Lion Swarm Optimization Algorithm (HSALSOA) combined with Modified Elliptic Curve Cryptography (ECC) is introduced to support safer sharing of data on wireless sensor networks (WSNs). The method suggests solutions for saving energy, managing crowded networks and ensuring dependable transmission in 50 sensor nodes. HSALSOA is applied to choose the best forwarder node using energy (E), delay (d) and fitness values as selection criteria. Results from the simulation show that the use of isocrete-path offers a significant improvement: The delay is now 0.02 sec, compared to ACO (0.11 sec) and throughput has improved to 0.88. We see that network lifetime reached an improved 6.8 units, compared to 0.8 units achieved by ACO. Because of multipath routing, devices avoid channel overloads and because of ECC, encryption errors are prevented. CBC with XOR padding is used in ECC-based signcryption with symmetric key encapsulation, increasing both confidentiality and data authenticity. As a result, HSALSO-ECC enhances both function and safety in real-time WSN projects centered on supporting agriculture, showing that it is a useful answer for IoT networks.
Privacy-Preserving NLP: Federated Learning for On-Device Language Model Fine-Tuning Ishta Rani, Laxmi Vanam, Dipesh Jagdish Kashiv, Amit Ojha, Arpit Jain, Savya Sachi 2025 International Conference on Sustainability Innovation and Technology Icsit 2025, 2025 Privacy-preserving natural language processing (NLP) is important for maintaining data confidentiality in applications with sensitive information. We investigate the use of federated learning (FL) to support on-device fine-tuning of language models (LMs), allowing collaborative model improvements without requiring raw user data centralization. In this paradigm, pre-trained language models are shipped to user devices and locally fine-tuned on private texts for taskspecific adaptation. Rather than sharing raw data stored on the local devices, devices compute model updates, and then only the encrypted gradients or model parameters are communicated to the centralized server. The server accumulates these updates to compose a global model, which is re-distributed to devices repeatedly in an iterative sequence, in an attempt to improve the performance of the network step by step. This federated framework tackles privacy issues by leaving sensitive data on the device and exploiting diverse linguistic distributions for strong personalization and generalization. Experiments on language modeling and text classification benchmarks verify its effectiveness by showing that FL-based fine-tuning provides similar performance as the traditional centralized fine-tuning while only experiencing a small accuracy decrement. Encrypted search, moreover, was enriched with features such as differential privacy and secure aggregation to provide further protection against leaks of sensitive information. On-device AI personalization at scale is based on the privacy-preserving NLP pipeline, and the potential to develop safe, user-centered language applications in the fields of healthcare, finance, and mobile communication, etc.
Improving the Performance of Convolutional Neural Networks for Image Classification Laxmi Vanam, Savya Sachi, Hina Gandhi, Sanjeev Kumar Shah, Madhusudana Kamballi, Punit Goel 2025 2nd International Conference on New Frontiers in Communication Automation Management and Security Iccams 2025, 2025 Image classification is a crucial part of computer vision that has applications in security, autonomous systems and healthcare. Enhanced classification accuracy has been achieved by Convolutional Neural Networks (CNN) by autonomous learning of hierarchical feature representations in images. The study's primary objective is to advance and improve the architecture of convolutional neural networks in order to achieve higher level accuracy in image classification. Multiple vital measures are used including augmenting the data, normalization in batches, adaptive learning rate scheduling and dropout to avoid overfitting and enhance the method's generalizability. In classification studies carried out on typical datasets like CIFAR-10 and MNIST, the results show that the method outperforms conventional neural network algorithms. In order to attain better accuracy, the research stresses the necessity for deeper approaches and more sophisticated training process. The suggested method offers a scalable and reliable answer to many practical image classification problems making a significant contribution to deep learning and computer vision
Enhancing Software Deployment Pipelines with the Application of ML Techniques Sourabh Sanghi, Shrinidhi Kota Shreeshapuranik, Na Kyung Kim, Devanand Ramachandran, Savya Sachi, Niharika Singh 2025 2nd International Conference on New Frontiers in Communication Automation Management and Security Iccams 2025, 2025 Continuous integration and deployment (CI/CD) pipelines, which are infamous for their slowness, frequent failures, and protracted build times, are a crucial part of the modern software development process. This study introduces a framework that uses machine learning (ML) to predictably boost pipeline performance. A Support Vector Machine technique for pipeline failure prediction will be developed as part of this effort. The method will also optimize resource allocation to decrease build times and provide dynamic framework for continuous development of CI/CD pipelines. This study assumes a comprehensive review of the literature and introduces a new approach based on the support vector machine (SVM) technology. The results demonstrate that the framework optimizes memory utilization and CPU, reduces build time by 33% and declines test success and failure rates by 60%. It also dramatically lowers resource consumption. The results showed that current DevOps approaches are able to tackle long-standing issues with CI/CD since they are intelligently scalable. This study laid the framework for future work in software engineering workflows including adaptive systems by introducing the idea of ML to the CI/CD process with the goal of improving pipeline efficacy and reliability.
Multimodal Emotion Recognition in Conversational AI Using Speech and Text Fusion Sunjhla Handa, Ramesh Kumar, Shrinidhi Kota Shreeshapuranik, Sourabh Sanghi, Arpit Jain, Savya Sachi 2025 International Conference on Sustainability Innovation and Technology Icsit 2025, 2025 One such factor is multimodal affect recognition to enhance conversational AI, since systems may be better contextually able to gauge the emotion of the user when presented with multimodal information. This paper offers a multimodal emotion recognition in conversational artificial intelligence based on deep learning involving the fusion of speech and text modalities. Model Our suggested model uses the state-of-the-art pre-trained language models (such as BERT) to extract textual features of the user turns, and CNNs or transformer-based encoders to process the speech signals, extracting the prosodic and paralinguistic features. Emotion classification is performed in the joint representation to make the system exploit the strengths of the speech and text that are mutually complementary. The model is proved to be efficient because the large-scale experiments are conducted on the IEMOCAP and MELD test sets. Results show that our model has also been found to be superior to the unimodal baselines when it comes to the recognition of emotions particularly in the challenging conversation environments. Even better, the suggested architecture is not merely solving the issue of modality imbalance and fusion schemes, but also facilitates real-time inference, which is encouraging when used in the framework of real-life conversational agents. It is a multimodal methodology as it enables dialogue systems grounded in empathy and emotional intelligence and leads to improved user satisfaction and interaction with AI worldview.
IoT Based Health Monitoring System with AI Powered Disease Prediction Sarvesh Kumar Gupta, Vamsi Krishna Gottipati, Savya Sachi, Ashish Gupta, Sanjeev Kumar Shah, Om Goel 2025 2nd International Conference on New Frontiers in Communication Automation Management and Security Iccams 2025, 2025 The COVID-19 pandemic wreaked havoc on many people's lives and livelihoods, as well as on many parts of the economy. As the number of patients increased at an exponential rate, the already-stressed healthcare industry encountered new problems. As a result, virtual consultation, telemedicine, and teleconsultation grew in popularity to adhere to social distance standards. This paper aims to grow a new system that uses the most effective machine learning (ML) algorithm to predict a patient's disease based on their signs, recordings of audio, available medical records, and other health histories. The goal is to address the urgent need for more "remote" discussions in the "post-COVID" era, and the Internet of Things (IoT) shows great promise in this regard. Sensors that are based on Arduino and ESP8266 may also be used to measure some of the indications, like fever and low oxygen levels in the blood, among others. It then offers for the suitable treatment and diagnosis of the condition based on its database, which is continually updated and may be designed as a platform that is either based on an application or a website.
Data Lake Validation Strategies: Ensuring Quality in Data Warehousing Pipelines Savya Sachi, Ravi Kiran Pagidi, Shilesh Karunakaran, Sarvesh Kumar Gupta, Suraj Dharmapuram, Om Goel 2025 International Conference on Intelligent and Secure Engineering Solutions Cises 2025, 2025 The most recent trend is storing vast quantities of raw and unstructured data. Nonetheless, this still places enormous challenges on the quality and reliability of the data in these lakes to support data warehousing pipelines. Any inaccurate, incomplete, or outdated data used in business analytics can produce incorrect insights and decisions, resulting in potential losses for organizations. So, it is essential to enforce data quality and make sure data warehousing pipelines are operating to their best by introducing data lake validation best practices. A critical approach that is often used is data profiling, where we analyze the data to detect patterns, relationships, and anomalies. This is the main reason why it is always a good practice to review the data first since this enables data engineers to gather information about the nature of the data and see its quality. Moreover, data can be cleaned by using techniques to find and correct errors in data, such as missing data or inconsistencies. The reason is that this step allows for the cleansing of data and ultimately ensures that downstream data processing can be done accurately.
Data Backup and Recovery in Cloud Storage with Data Security 14th International Conference on Advances in Computing Control and Telecommunication Technologies Act 2023, 2023
A Technique for Improving High-Speed Dynamic Networks using IoT 14th International Conference on Advances in Computing Control and Telecommunication Technologies Act 2023, 2023
Twitter Fake News Detection by Using Xlnet Model Senthil Athithan, Savya Sachi, Ajay Kumar Singh, Arpit Jain, Divya, Yogesh Kumar Sharma Proceedings International Conference on Technological Advancements in Computational Sciences Ictacs 2023, 2023
Fake Currency Detection using Ensemble Learning Ashok Kumar, Savya Sachi, Anshoo Bhatia, Pravesh Belwal, Santosh Kumar, Vivek Bhatnagar 2023 10th IEEE Uttar Pradesh Section International Conference on Electrical Electronics and Computer Engineering Upcon 2023, 2023
AI - Inspired Algorithms for the Diagnosis of Diseases in Cotton Plant Neetu, Babita Rani Radwal, Savya Sachi, Santosh Kumar, Arpit Jain, Satendra Kumar 2023 10th IEEE Uttar Pradesh Section International Conference on Electrical Electronics and Computer Engineering Upcon 2023, 2023
Highly Secured Image Fringe Steganography Savya Sachi, Santosh Kumar, Ashok Kumar, Vivek Bhatnagar, Abhishek Jain, Vibhoo Sharma Proceedings International Conference on Technological Advancements in Computational Sciences Ictacs 2023, 2023
Copy and Move Forged Image Detection by Deep Learning Yogesh Kumar Sharma, Senthil Athithan, Savya Sachi, Ajay Kumar Singh, Arpit Jain, Suman Devi 2023 World Conference on Communication and Computing Wconf 2023, 2023
Optimized Protected Data Communication in WSN Using Fusion Pretended Tempering with Lion Swarm Optimization and Improved Elliptic Curve Cryptography S Sahu, AP Singh, B Seva, V Rathod, M Shamila, S Sachi 2025 IEEE 5th International Conference on ICT in Business Industry … , 2025 2025
An IoT-Based Traffic Management System Using a Mean Fitness-Oriented Dragonfly Algorithm for Traffic Flow Prediction. P Dayalan, PB Rao, R Kumar, N Nigam, V Midasala, S Sachi, V Roy Ingénierie des Systèmes d'Information 30 (11) , 2025 2025
Impact of Bihar Entrepreneurship Scheme on Migrant Worker Retention and Socio-Economic Development: A Comprehensive Analysis DSS Dr. Ritu Narayan, Vibhash Ranjan Journal of Asia Entrepreneurship and Sustainability 21 (03), 1-5 , 2025 2025
Multimodal Emotion Recognition in Conversational AI Using Speech and Text Fusion S Handa, R Kumar, SK Shreeshapuranik, S Sanghi, A Jain, S Sachi 2025 International Conference on Sustainability, Innovation & Technology … , 2025 2025
Bias Mitigation in Generative Chatbots Through Adversarial Debiasing P Kumar, SK Venugopal, S Sachi, S Handa, SK Gupta, A Jain 2025 International Conference on Sustainability, Innovation & Technology … , 2025 2025
Privacy-Preserving NLP: Federated Learning for On-Device Language Model Fine-Tuning I Rani, L Vanam, DJ Kashiv, A Ojha, A Jain, S Sachi 2025 International Conference on Sustainability, Innovation & Technology … , 2025 2025
Data Lake Validation Strategies: Ensuring Quality in Data Warehousing Pipelines S Sachi, RK Pagidi, S Karunakaran, SK Gupta, S Dharmapuram, O Goel 2025 International Conference on Intelligent and Secure Engineering … , 2025 2025
Enhancing Software Deployment Pipelines with the Application of ML Techniques S Sanghi, SK Shreeshapuranik, NK Kim, D Ramachandran, S Sachi, ... 2025 2nd International Conference on New Frontiers in Communication … , 2025 2025 Citations: 1
Improving the Performance of Convolutional Neural Networks for Image Classification L Vanam, S Sachi, H Gandhi, SK Shah, M Kamballi, P Goel 2025 2nd International Conference on New Frontiers in Communication … , 2025 2025
IoT Based Health Monitoring System with AI Powered Disease Prediction SK Gupta, VK Gottipati, S Sachi, A Gupta, SK Shah, O Goel 2025 2nd International Conference on New Frontiers in Communication … , 2025 2025 Citations: 1
An Operative Expectation of Parkinson’s Ailment Using a Hybrid Machine Learning and Artificial Intelligence Systems S Athithan, S Sachi, AK Singh SN Computer Science 5 (8), 1146 , 2024 2024 Citations: 1
Imagine Segmentation for Brain Tumor Prognosis A Chouksey, S Sachi, S Athitan, R Shrivastava, A Purohit, S Kumar 2024 IEEE 6th International Conference on Cybernetics, Cognition and Machine … , 2024 2024
Efficient Protocol Selection and Estimation in Wireless Sensor Networks. P Sinha, MM Alam, S Sachi, N Priya, V Kumar Journal of Computational Analysis & Applications 33 (2) , 2024 2024
MAKING USE OF MANUFACTURING PROCESS VARIATIONS: MACHINE LEARNING APPROACHES FOR EFFICIENT MEDICAL AND BIOLOGICAL STUDY-BASED IMAGE COMPRESSION AND LOSSLESS TRANSMISSION. S Sachi, R Ranjan, S Kumari, MM Alam Biochemical & Cellular Archives 24 (2) , 2024 2024
MACHINE LEARNING APPROACHES FOR PREDICTING PROTEINPROTEIN INTERACTIONS. S Kumar, S Sachi, RM Tugnayat Biochemical & Cellular Archives 24 (2) , 2024 2024
INTEGRATING BIG DATA ANALYTICS IN THE STUDY OF ANIMAL POPULATION DYNAMICS. R Kumar, RK Roshan, A Kumar, S Sachi, MM Alam Biochemical & Cellular Archives 24 (2) , 2024 2024
A SURVEY ON PROTEIN-TO-PROTEIN INTERACTION PREDICTION USING TRANSFER LEARNING. A Alam, S Sachi, S Kumar Biochemical & Cellular Archives 24 (2) , 2024 2024 Citations: 1
INTEGRATING COMPUTATIONAL METHODS IN PURE AND APPLIED ZOOLOGY: ENHANCING SPECIES CONSERVATION AND BEHAVIORAL STUDIES. S Sachi, S Kumar Biochemical & Cellular Archives 24 (2) , 2024 2024 Citations: 1
SYSTEMS BIOLOGY APPROACHES TO UNDERSTANDING GENE REGULATORY NETWORKS IN DEVELOPMENTAL PROCESSES FOR PREDICTING PROTEIN-PROTEIN INTERACTIONS. S Sachi, A Alam Biochemical & Cellular Archives 24 (2) , 2024 2024
Securing the Digital Commerce Spectrum and Cyber Security Strategies for Web, E-commerce, M-commerce, and E-mail Security. R Pachlor, R Mohanraj, K Sharada, S Sachi, K Neelima, P Ramadevi Journal of Cybersecurity & Information Management 14 (1) , 2024 2024 Citations: 1
MOST CITED SCHOLAR PUBLICATIONS
AI-Inspired Algorithms for the Diagnosis of Diseases in Cotton Plant BR Radwal, S Sachi, S Kumar, A Jain, S Kumar 2023 10th IEEE Uttar Pradesh Section International Conference on Electrical … , 2023 2023 Citations: 54
An efficient Intra-Inter pixel encryption scheme to secure healthcare images for an IoT environment S Dash, S Padhy, SA Devi, S Sachi, KAK Patro Expert systems with applications 231, 120622 , 2023 2023 Citations: 43
HDL environment for the synthesis of 2-dimensional and 3-dimensional network on chip mesh router architecture S Kumari, K Rajput, G Singh, A Jain, S Sachi, M Manwal 2024 International conference on communication, computer sciences and … , 2024 2024 Citations: 19
Fire detection using image processing technique A Jain, YK Sharma, S Sachi, S Athithan, AK Singh 2023 3rd International Conference on Technological Advancements in … , 2023 2023 Citations: 19
A Discrete-Time Image Hiding Algorithm Transform Using Wavelet and SHA-512 S Kumar, S Sachi, A Kumar, A Jain, MSR Prasad 2023 3rd International Conference on Technological Advancements in … , 2023 2023 Citations: 19
Multi objective optimization based land cover classification using NSGA-II AK Singh, A Jain, YK Sharma, S Athithan, S Sachi 2023 6th International Conference on Contemporary Computing and Informatics … , 2023 2023 Citations: 19
Implementation of ABC & WOA-based security defense mechanism for distributed denial of service attacks S Devi, YK Sharma, S Athithan, S Sachi, AK Singh, A Jain 2023 6th International Conference on Contemporary Computing and Informatics … , 2023 2023 Citations: 19
Twitter fake news detection by using Xlnet model S Athithan, S Sachi, AK Singh, A Jain, YK Sharma 2023 3rd International Conference on Technological Advancements in … , 2023 2023 Citations: 15
Fog Restoration in Hazy Images using Deep Transfer Learning NK Agrawal, V Sharma, P Singh, S Sachi, A Jain, MM Alam 2023 International Conference on Smart Devices (ICSD), 1-5 , 2024 2024 Citations: 13
Hate speech detection using the GPT-2 and natural language processing S Sachi, AK Singh, A Jain, S Devi, YK Sharma, S Athithan 2023 Intelligent Methods, Systems, and Applications (IMSA), 276-280 , 2023 2023 Citations: 12
Copy and Move Forged Image Detection by Deep Learning YK Sharma, S Athithan, S Sachi, AK Singh, A Jain, S Devi 2023 World Conference on Communication & Computing (WCONF), 1-6 , 2023 2023 Citations: 12
Hy_PSO: Hybrid Algorithm for Lung Cancer Diagnosis and Prognosis S Sachi, J Jain, A Jain, UK Patel, A Bhatnagar, A Jain 2023 International Conference on Smart Devices (ICSD), 1-5 , 2024 2024 Citations: 11
Icme for Process Scale‐Up: Importance of Vertical and Horizontal Integration of Models G Tennyson, R Shukla, S Mangal, S Sachi, AK Singh Proceedings of the 3rd World Congress on Integrated Computational Materials … , 2015 2015 Citations: 11
Ultrasound-Based Ovarian Cysts Detection with Improved Machine-Learning Techniques and Stage Classification Using Enhanced Classifiers AKS Senthil Athithan, Savya Sachi S N COMPUTER SCIENCE 4 (5), 1-11 , 2023 2023 Citations: 9
A framework for morphological operations using counter harmonic mean S Sachi, D Ganesh, L Bhagyalakshmi, R Tiwari, SK Suman, SP Manikanta, ... Proceedings on Engineering Sciences 6 (4), 1531-1540 , 2024 2024 Citations: 6
Network on chip for 2d mesh toplological structure in hdl enviornment S Kumar, A Kumar, V Rana, V Sharma, V Bhatnagar, S Sachi 2023 10th IEEE Uttar Pradesh Section International Conference on Electrical … , 2023 2023 Citations: 6
Detection of Online Humiliation Through Social Media Platforms Using AI Inspired Algorithms A Bhatia, A Kumar, A Kumar, S Sachi, S Kumar 2023 3rd International Conference on Technological Advancements in … , 2023 2023 Citations: 6
An analysis of election prediction using social media data network: a review S Nayeem, S Sachi, R Kumar Journal of Aeronautical Materials 43 (01), 290-298 , 2023 2023 Citations: 4
Classification of robots by environment and mechanism of interaction S Sachi, N Kumar International Journal of Advanced Research in Science, Communication and … , 2020 2020 Citations: 3
Software-Defined Network Frameworks: Security Issues and Use Cases M Kaur, V Jain, P Nand, N Rakesh CRC Press , 2024 2024 Citations: 2