Digitala Vetenskapliga Arkivet

Change search
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf
LLMs and RAG for FinancialData in the FinTech Domain
Umeå University, Faculty of Science and Technology, Department of Computing Science.
2025 (English)Independent thesis Advanced level (degree of Master (Two Years)), 20 credits / 30 HE creditsStudent thesis
Abstract [en]

Abstract. This thesis investigates the implementation and impact of Large Language Models (LLMs) and Retrieval Augmented Generation (RAG) in financial risk management. Through qualitative research methods and iterative prototype development in a financial institution, this thesis explores how these technologies can improve data accessibility and decision-making processes for risk management professionals. The findings reveal both the potential and challenges in the implementation of these technologies in financial environments. Although risk managers showed great enthusiasm and trust in the technology, successful implementation required substantial manual engineering effort in data normalization, semantic modeling, and query generation. Key challenges included handling ambiguous natural language queries and maintaining accuracy in financial calculations. The thesis demonstrates that while LLMs and RAG can improve data accessibility, their effective deployment requires careful attention to domain-specific requirements and human factors.The research contributes to the growing body of knowledge on practical applications of AI in financial services by providing insight into implementation challenges, user acceptance, and system design requirements. These findings have important implications for financial institutions considering similar implementations and suggest directions for future research to improve the reliability and effectiveness of AI-assisted financial risk management tools.

Place, publisher, year, edition, pages
2025.
Series
UMNAD ; 1528
National Category
Computer and Information Sciences
Identifiers
URN: urn:nbn:se:umu:diva-236103OAI: oai:DiVA.org:umu-236103DiVA, id: diva2:1942374
External cooperation
Clear Street
Educational program
Master of Science Programme in Computing Science and Engineering
Supervisors
Examiners
Available from: 2025-03-05 Created: 2025-03-04 Last updated: 2025-03-05Bibliographically approved

Open Access in DiVA

fulltext(288 kB)225 downloads
File information
File name FULLTEXT01.pdfFile size 288 kBChecksum SHA-512
66047a58277b698c1aa787a76e071238f1f936a87e4262add7fae17c4d63af266596aa8d8c824f77b0376f7d81a63e5d974987f3ca738c738bae2e02795e91a3
Type fulltextMimetype application/pdf

By organisation
Department of Computing Science
Computer and Information Sciences

Search outside of DiVA

GoogleGoogle Scholar
Total: 225 downloads
The number of downloads is the sum of all downloads of full texts. It may include eg previous versions that are now no longer available

urn-nbn

Altmetric score

urn-nbn
Total: 503 hits
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf