Benchmarking AWS Bedrock Generative AI Models with RAG for Analyzing Fraudulent Transactions

Vekariya, Hunaid (2024) Benchmarking AWS Bedrock Generative AI Models with RAG for Analyzing Fraudulent Transactions. Masters thesis, Dublin, National College of Ireland.

Preview	PDF (Master of Science) Download (1MB) | Preview
Preview	PDF (Configuration Manual) Download (2MB) | Preview

Abstract

The proliferation of AI and Large Language Models (LLMs) are continuously evolving, with new models developed and tuned at high rate. However, they still facing limitations in providing domain-specific responses, particularly when dealing with fraudulent transactional data and financial knowledge. This challenge can be addressed through retrieval-augmented generation (RAG) systems deployed on serverless opensearch platform, which integrate vector databases to enhance model performance. This research benchmarks two advanced models from aws bedrock viz, Titan G1 premier and Mistral AI large. Comparing their performance based on execution time, token generation efficiency, comprehensiveness, and their compatibility with serverless event-driven architecture as these models are note evaluated on analyzing fraud activities. By analyzing how AI models can detect and mitigate fraudulent transactions, this study highlights the potential of combining LLMs with RAG systems to improve fraud detection and prevention. The evaluation findings are presented where Titan model provides more accurate and faster results, offering valuable insights and paving the way for further research opportunities.

Item Type:	Thesis (Masters)
Supervisors:	Name Email Kazmi, Aqeel UNSPECIFIED
Uncontrolled Keywords:	Prompt Engineering; AWS Bedrock; GenAI; Amazon Titan; Mistral AI; Credit Card Transaction; Opensearch Serverless
Subjects:	Q Science > QA Mathematics > Electronic computers. Computer science T Technology > T Technology (General) > Information Technology > Electronic computers. Computer science Q Science > QH Natural history > QH301 Biology > Methods of research. Technique. Experimental biology > Data processing. Bioinformatics > Artificial intelligence Q Science > Q Science (General) > Self-organizing systems. Conscious automata > Artificial intelligence T Technology > T Technology (General) > Information Technology > Cloud computing H Social Sciences > HG Finance > Credit. Debt. Loans.
Divisions:	School of Computing > Master of Science in Cloud Computing
Depositing User:	Ciara O'Brien
Date Deposited:	17 Jul 2025 12:59
Last Modified:	17 Jul 2025 12:59
URI:	https://norma.ncirl.ie/id/eprint/8163

Actions (login required)

View Item