Agentic Offer Creation Workflow

This thesis develops and evaluates an AI assistant that automates the sales offer creation process at Your Sidekicks AG. The system demonstrably improves efficiency and quality, leading to significant time and cost savings.

Omar Rahiel, 2025

Art der Arbeit Bachelor Thesis
Auftraggebende Your Sidekicks AG
Betreuende Dozierende Pustulka, Elzbieta
Views: 8
As the digital consulting firm Your Sidekicks AG strategically shifts from standard marketing services to more complex digital consulting offerings, its manual offer creation process has become a bottleneck. Creating tailored proposals for these new, sophisticated projects is inefficient and inconsistent, lacking the operational scalability needed to support the company's strategic evolution and growing client demand.
Following the Design Science Research (DSR) approach, an analysis of the manual "As-Is" workflow identified key bottlenecks, leading to the development of the "Agentic Offer Creation Workflow." This LLM-based system uses Retrieval-Augmented Generation (RAG) to ground its outputs in company knowledge for factual accuracy. A crucial "human-in-the-loop" step secures consultant control and trust, before the workflow concludes by automatically creating a final quote directly in the company's accounting system.
The result is a validated prototype that significantly improves the efficiency, quality, and scalability of offer creation. - Efficiency: A quantitative analysis revealed a reduction in the time required for a complex offer by 81.25%. This corresponds to a projected annual saving of over 260 working hours. - Quality: A qualitative evaluation, including a blind comparative review by internal experts, showed that the AI-generated offers surpassed the quality of manually created ones (average score of 9.41 vs. 9.16). In the field test, all AI-generated offers passed the internal quality assurance process. - Operational Scalability: The artifact dramatically enhances operational scalability. The end-to-end integration with Bexio is the crucial final step that makes these efficiency gains fully operational across the entire workflow, allowing the firm to handle a greater volume of proposals without a proportional increase in consultant workload. The thesis provides a functional IT prototype and concrete, actionable recommendations for Your Sidekicks AG.
Studiengang: Business Information Technology (Bachelor)
Keywords Agentic AI, Process Optimization, Large Language Models (LLMs), Retrieval-Augmented Generation (RAG), Human-in-the-Loop, Proof-of-Concept (PoC), Design Science Research
Vertraulichkeit: öffentlich
Art der Arbeit
Bachelor Thesis
Auftraggebende
Your Sidekicks AG, Basel
Autorinnen und Autoren
Omar Rahiel
Betreuende Dozierende
Pustulka, Elzbieta
Publikationsjahr
2025
Sprache der Arbeit
Englisch
Vertraulichkeit
öffentlich
Studiengang
Business Information Technology (Bachelor)
Standort Studiengang
Basel
Keywords
Agentic AI, Process Optimization, Large Language Models (LLMs), Retrieval-Augmented Generation (RAG), Human-in-the-Loop, Proof-of-Concept (PoC), Design Science Research