Authors:
K. Balaji, S. Silvia Priscila, B. M. Praveen

Addresses:
Institute of Computer Science and Information Science, Srinivas University, Dakshina Kannada, Karnataka, India. Department of Computer Science and Information Technology, School of Computing Sciences, Vels Institute of Science, Technology and Advanced Studies, Chennai, Tamil Nadu, India. Department of Computer Science, Bharath Institute of Higher Education and Research, Chennai, Tamil Nadu, India. Institute of Engineering and Technology, Srinivas University, Dakshina Kannada, Karnataka, India.

Abstract:

The digital transformation of enterprises is occurring so quickly that attack surfaces are now much larger, and intrusion detection and cyber risk management in real time are becoming more complex. Traditional signature-based and rule-driven security systems are unable to handle high-volume, high-velocity network traffic and evolving attack patterns, leading to delayed attack detection and greater breach impact. The goal of this research is to implement and test the performance of real-time deep learning engines for detecting sophisticated intrusions while enabling enterprise-level cyber risk management. The current framework proposes combining deep neural architectures (i.e., convolutional neural networks to learn spatial features and recurrent models to learn temporal dependencies) within the streaming analytics pipeline. The system is distributed to provide low-latency inference for continuous network flows. Experimental evaluation on large-scale benchmark and enterprise-mimicking datasets shows that the proposed engine achieves 98.6% detection accuracy, reduces false-positive rates by 32% compared to traditional IDS solutions, and maintains an average detection latency of less than 120 ms under peak loads. These results indicate that real-time deep learning engines can significantly improve enterprise intrusion detection capabilities, achieving both high accuracy and operational scalability. The conclusion of the given study is that deep learning-driven, real-time IDS frameworks are a viable foundation for next-generation cyber risk management systems in enterprise environments.

Keywords: Enterprise Security; Network Security; Deep Learning (DL); Intrusion Detection Systems; Digital Transformation; Deep Learning-Driven; Cyber Risk Management.

Received on: 04/02/2025, Revised on: 13/04/2025, Accepted on: 27/06/2025, Published on: 03/01/2026

DOI: 10.69888/FTSIN.2026.000601

FMDB Transactions on Sustainable Intelligent Networks, 2026 Vol. 3 No. 1, Pages: 1–14

• Views : 55
• Downloads : 11

Download PDF