Deepak Kaul

Follower

Following

Public Views

With 20 years of experience in the software industry, I specialize in secure, scalable architectures for enterprise applications, with expertise in microservices, cloud technologies (AWS/Azure), API security, and distributed systems. My contributions span prominent organizations like Marriott, USAA, and Charter Communications, where I focused on innovative solutions in Spring Boot, RESTful APIs, and cloud-native applications.An accomplished researcher, I have published 10 papers across various journals. My interests lie in AI/ML, cybersecurity, multi-cloud environments, and microservices, with a focus on solving complex challenges in hotel reservation systems and e-commerce platforms.I am currently actively working on a patent for an Intelligent Data Privacy and Compliance Platform, "Guardian Sense," aimed at revolutionizing enterprise-level data security and cloud compliance.

less

Interests

Uploads

Papers by Deepak Kaul

AI-DRIVEN DYNAMIC UPSELL IN HOTEL RESERVATION SYSTEMS BASED ON CYBERSECURITY RISK SCORES

International Journal of Computer Engineering and Technology (IJCET), 2021

In the era of digital transformation, hotels and travel companies are increasingly targeted by cy... more In the era of digital transformation, hotels and travel companies are increasingly targeted by cybersecurity threats such as phishing attacks, data breaches, and ransomware. These threats compromise the integrity of booking systems and customer data. This paper proposes a novel approach to dynamic upselling in hotel reservation systems by leveraging AI to assess real-time cybersecurity risk scores of customer accounts. Based on the risk assessment, the system will offer upsell options such as identity theft insurance, secure payment options, and other security-related add-ons, ensuring customers' peace of mind while capitalizing on business opportunities. The key contribution of this research lies in integrating AI-based cybersecurity analysis with dynamic upselling strategies, which is underexplored in current literature. We conduct a comprehensive review of the interplay between AI-driven cybersecurity and upselling in digital commerce, proposing a model that responds to real-time risks to safeguard customer data while enhancing revenue.

Download

AI-DRIVEN DECENTRALIZED AUTHENTICATION SYSTEM USING HOMOMORPHIC ENCRYPTION

IAEME, 2022

In today's highly digital environment, centralized authentication systems are susceptible to sing... more In today's highly digital environment, centralized authentication systems are susceptible to single points of failure, exposing sensitive data to cyberattacks and data breaches. This paper presents a novel AI-driven decentralized authentication system utilizing homomorphic encryption, designed to enhance data privacy and security. Unlike conventional decentralized systems, the integration of AI for identity verification, combined with homomorphic encryption, allows for operations on encrypted data without decryption, significantly reducing the risk of data exposure. This method can be applied to sectors where data privacy is paramount, such as finance and healthcare. Our research demonstrates that the proposed system ensures robust, secure, and scalable identity verification while maintaining data confidentiality, even in compromised environments. The findings are supported by a detailed comparison with traditional centralized authentication models and other decentralized methods. The results indicate a significant reduction in attack vectors and improved privacy outcomes.

Download

Adaptive Neuro -Symbolic AI for Autonomous Decision -Making in High -Stakes Mission -Critical Systems

International Journal of Science and Research, 2024

Autonomous decision-making in mission critical domains like aerospace, healthcare, defense, and t... more Autonomous decision-making in mission critical domains like aerospace, healthcare, defense, and transport systems, is essential to ensure the resilience of a system and to enable rapid responses (Chen & Lee, 2022). Scientists and academics are of the view that the conventional AI algorithms often fail to work effectively owing to the need for adaptability in these scenarios and the fact that their performance can be understood (Smith & Jones, 2023). Neuro-symbolic AI, a hybrid paradigm that couples neural networks based pattern recognition with the classical and logical reasoning of symbolic AI, is another (Garcez et al., 2020). This study investigates the construction and refinement of an intelligent Decorator such that it can effectively support and escalate the treatment of decision making inaccuracies due to the variations, shifts and lack of clear direction in the missions. To safely co-locate actual tests, most volatile deigned burglary patterns were used while the advertisements in the background offered sweets or sometimes indicated that the alarm would ring in five minutes. This paper will illustrate that neuro-symbolic AI is superior to traditional models for two aspects of discrepancy: that of precision and that of model interpretability.

Download

AI-DRIVEN DYNAMIC UPSELL IN HOTEL RESERVATION SYSTEMS BASED ON CYBERSECURITY RISK SCORES

International Journal of Computer Engineering and Technology (IJCET), 2021

Download

AI-DRIVEN DECENTRALIZED AUTHENTICATION SYSTEM USING HOMOMORPHIC ENCRYPTION

IAEME, 2022

Download

Adaptive Neuro -Symbolic AI for Autonomous Decision -Making in High -Stakes Mission -Critical Systems

International Journal of Science and Research, 2024

Download

Deepak Kaul

Uploads

Papers by Deepak Kaul

Log In