kth.sePublications
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
Generative AI for Sales: A Study on the Potential of Retrieval Augmented Generation for Response Automation in the Request for Proposal Process in Telecommunications
KTH, School of Electrical Engineering and Computer Science (EECS).
KTH, School of Electrical Engineering and Computer Science (EECS).
2024 (English)Independent thesis Basic level (degree of Bachelor), 10 credits / 15 HE creditsStudent thesisAlternative title
Generativ AI för Försäljning : En Studie om Potentialen om Retrieval Augmented Generation för Svars Automation i Request for Proposal Processen i Telekommunikation (Swedish)
Abstract [en]

The Request for Proposal (RFP) is a sales document containing different requirements with the goal to solicit a supplier. The RFP process is critical in telecommunications sales, requiring efficient and accurate responses to secure projects. This paper aims to provide a proof of concept (POC) for a Generative AI (GenAI) solution for RFP response automation. Retrieval Augmented Generation (RAG) was used and is a framework that enhances GenAI models by retrieving and augmenting domain-specific information from an external knowledge base. This study’s methodology mainly involved implementing and evaluating RAG with different chunking variations, as well as performing work system analysis on the RFP response creation process to recommend a ’to-be’ system. Results indicate that simple technical requirements had more accurate responses with a smaller chunking strategy, while a bigger chunking strategy was suitable for advanced technical requirements. Overall, the parent and child chunking strategy performed best for all requirements. This study concluded that a POC can be created and contributes valuable insights on the implementation of RAG in a business process, emphasizing its potential benefits and risks.

Abstract [sv]

Request for Proposal (RFP) är ett försäljningsdokument som innehåller olika krav med målet att värva en leverantör. RFP-processen är avgörande vid telekommunikationsförsäljning och kräver effektiva och korrekta svar för att säkra projekt. Detta dokument syftar till att tillhandahålla ett proof of concept (POC) för en Generativ AI (GenAI)-lösning för RFP-svarsautomatisering. Retrieval Augmented Generation (RAG) användes och är ett ramverk som förbättrar GenAI-modeller genom att hämta och utöka domänspecifik information från en extern kunskapsbas. Denna studies metodik involverade huvudsakligen implementering och utvärdering av RAG med olika chunking-variationer, samt att utföra arbetssystemanalyser på RFP-svarsskapandeprocessen för att rekommendera ett ”to-be”-system. Resultaten indikerar att enkla tekniska krav hade mer exakta svar med en mindre chunking strategi, medan en större chunking strategi var lämplig för avancerade tekniska krav. Sammantaget fungerade strategin ’parent and child chunking’ . . bäst för alla krav. Denna studie drog slutsatsen att en POC kan skapas och bidrar med värdefulla insikter om implementeringen av RAG i en affärsprocess, och betonar dess potentiella fördelar och risker.

Place, publisher, year, edition, pages
2024. , p. 14
Series
TRITA-EECS-EX ; 2024:271
Keywords [en]
GenAI, RAG, RFP, telecommunications, WST
National Category
Computer and Information Sciences
Identifiers
URN: urn:nbn:se:kth:diva-350770OAI: oai:DiVA.org:kth-350770DiVA, id: diva2:1884900
Supervisors
Examiners
Available from: 2024-08-13 Created: 2024-07-18 Last updated: 2024-08-13Bibliographically approved

Open Access in DiVA

fulltext(820 kB)585 downloads
File information
File name FULLTEXT01.pdfFile size 820 kBChecksum SHA-512
04b599fb2a7d251e7c8d40abdd89f6228c3421022bcfc1b66a1fb82ca98ce1082229eabf6fe53cf043b5ea142527995b6984125ae6b514a82001abc3956a118d
Type fulltextMimetype application/pdf

By organisation
School of Electrical Engineering and Computer Science (EECS)
Computer and Information Sciences

Search outside of DiVA

GoogleGoogle Scholar
Total: 585 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: 891 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