AI-Powered Mobile Application for Automated Product Recognition, Description Generation, and Price Estimation
Amosun, Olumide (2025)
2025
All rights reserved. This publication is copyrighted. You may download, display and print it for Your own personal use. Commercial use is prohibited.
Julkaisun pysyvä osoite on
https://urn.fi/URN:NBN:fi:amk-2025060520723
https://urn.fi/URN:NBN:fi:amk-2025060520723
Tiivistelmä
This thesis explored the development of a mobile application that uses artificial intelligence (AI) to simplify and automate uploading and categorizing products for store owners, E-commerce platforms, inventory managers, as well as customers. The app allows users to capture or upload product images, which are then analyzed by AI algorithms. These algorithms identify the product category, generate appropriate titles and descriptions, and suggest estimated prices based on current market trends.
The mobile application was built using Flutter, and the backend utilized Google's Firebase and Python AI microservices. This work provides a proof of concept for the successful use of a cloud-based AI platform and mobile development as an Enabler for increasing operational efficiency in business within a digital Commerce ecosystem. The app's tech stack includes Amazon Rekognition, a deep learning-based image analysis service, on which object detection and classification were performed without needing to train custom image models. NLP techniques were used for generating consistent and contextually correct product titles and descriptions. To create price recommendations, real-time data retrieval from market sources was implemented.
The mobile application was built using Flutter, and the backend utilized Google's Firebase and Python AI microservices. This work provides a proof of concept for the successful use of a cloud-based AI platform and mobile development as an Enabler for increasing operational efficiency in business within a digital Commerce ecosystem. The app's tech stack includes Amazon Rekognition, a deep learning-based image analysis service, on which object detection and classification were performed without needing to train custom image models. NLP techniques were used for generating consistent and contextually correct product titles and descriptions. To create price recommendations, real-time data retrieval from market sources was implemented.