This dissertation explores how recent advancements in artificial intelligence can revolutionize After-Sales services by improving efficiency, customer satisfaction, and key performance metrics. Despite extensive research on AI in business, its integration into After-Sales services remains largely unexplored. This study bridges that gap by analyzing the feasibility, challenges, and benefits of AI-driven solutions, with a focus on the luxury fashion industry. It investigates two core implementations: Defect Detection using Computer Vision and Sentiment Analysis through Large Language Models. A qualitative approach integrates academic literature, industry case studies, and interviews with a luxury fashion brand and AI providers. The research identifies inefficiencies in After-Sales operations, such as long lead times, fragmented communication, and limited in-store expertise. To address these challenges, the Defect Detection model introduces an AI-powered system for automating product inspection, detecting defects, and providing real-time repair cost and timeline estimates. Meanwhile, Sentiment Analysis enables brands to monitor customer feedback, identify pain points and adjust After-Sales strategies accordingly, reinforcing customer loyalty and brand perception. To ensure a balance between technological feasibility and financial sustainability, a cost estimation analysis was conducted, comparing different AI implementation models: pre-trained, fine-tuned, and creation from scratch. Finally, the study applies the After Sales Performance Measurement Framework to assess AI’s impact on operational, process and strategic levels. Results indicate that AI-driven enhancements improve responsiveness, efficiency and customer alignment, positioning AI as a key driver of competitive advantage in the luxury fashion sector.
Questa tesi esplora come i recenti progressi nell'intelligenza artificiale possano rivoluzionare i servizi After-Sales migliorando l'efficienza, la soddisfazione del cliente e i principali indicatori di performance. Nonostante l'ampia ricerca sull'uso dell'IA nel settore aziendale, la sua integrazione nei servizi After-Sales rimane in gran parte inesplorata. Questo studio colma tale lacuna analizzando la fattibilità, le sfide ed i benefici delle soluzioni basate sull'IA, con un focus specifico sull'industria della moda di lusso. Lo studio esamina due implementazioni principali: Defect Detection attraverso modelli diComputer Vision e Sentiment Analysis attraverso l'uso di Large Language Models. L’approccio qualitativo adottato integra letteratura accademica, studi di caso nel settore e interviste con un marchio di moda di lusso e fornitori di soluzioni AI. La ricerca identifica le principali inefficienze nelle operazioni di post-vendita, come i lunghi tempi di attesa, la comunicazione frammentata e le limitate competenze tecniche del personale in negozio. Per affrontare queste Sfide, il modello di Defect Detection introduce un sistema basato sull’IA per automatizzare l’ispezione dei prodotti, rilevare difetti e fornire stime in tempo reale dei costi e i tempi di riparazione. Parallelamente, la Sentiment Analysis consente ai brand di monitorare il feedback dei clienti, identificare i punti piu critici e adattare di conseguenza le strategie post vendita, rafforzando la fedeltà dei clienti e la percezione del marchio. Per garantire un equilibrio tra fattibilità tecnologica e sostenibilità finanziaria, è stata condotta un'analisi di stima dei costi confrontando diversi modelli di implementazione dell’IA: pre-addestrati, fine-tuned e sviluppati da zero. Infine, lo studio applica il modello di Misurazione delle Performance After Sales per valutare l’impatto dell’IA a livello operativo, di processo e strategico. I risultati indicano che le soluzioni basate sull’IA migliorano l’efficienza e l’allineamento con le esigenze dei clienti, posizionando l’intelligenza artificiale come un fattore chiave di vantaggio competitivo nel settore della moda di lusso.
Artificial Intelligence for after sales services: a case of luxury fashion industry
Alfarano, Agnese
2024/2025
Abstract
This dissertation explores how recent advancements in artificial intelligence can revolutionize After-Sales services by improving efficiency, customer satisfaction, and key performance metrics. Despite extensive research on AI in business, its integration into After-Sales services remains largely unexplored. This study bridges that gap by analyzing the feasibility, challenges, and benefits of AI-driven solutions, with a focus on the luxury fashion industry. It investigates two core implementations: Defect Detection using Computer Vision and Sentiment Analysis through Large Language Models. A qualitative approach integrates academic literature, industry case studies, and interviews with a luxury fashion brand and AI providers. The research identifies inefficiencies in After-Sales operations, such as long lead times, fragmented communication, and limited in-store expertise. To address these challenges, the Defect Detection model introduces an AI-powered system for automating product inspection, detecting defects, and providing real-time repair cost and timeline estimates. Meanwhile, Sentiment Analysis enables brands to monitor customer feedback, identify pain points and adjust After-Sales strategies accordingly, reinforcing customer loyalty and brand perception. To ensure a balance between technological feasibility and financial sustainability, a cost estimation analysis was conducted, comparing different AI implementation models: pre-trained, fine-tuned, and creation from scratch. Finally, the study applies the After Sales Performance Measurement Framework to assess AI’s impact on operational, process and strategic levels. Results indicate that AI-driven enhancements improve responsiveness, efficiency and customer alignment, positioning AI as a key driver of competitive advantage in the luxury fashion sector.File | Dimensione | Formato | |
---|---|---|---|
2025_04_Alfarano_Agnese.pdf
accessibile in internet solo dagli utenti autorizzati
Descrizione: Executive Summary + Thesis
Dimensione
3.24 MB
Formato
Adobe PDF
|
3.24 MB | Adobe PDF | Visualizza/Apri |
I documenti in POLITesi sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/10589/236440