Implementing an AI system to assessuser story quality

Description : Agileteams break down user needs into smallfeatures or tasksthat can be implemented and delivered from few hours to few days. The Product Owner actively prioritizes these user stories in terms of user value, risk, and business value to significantly increase the value delivered by a team within the first few sprints. Those user stories delivered to Dev team must be with a good quality to ensure Highest Value Delivery, Fosters Collaboration, Boost Transparency, Minimalnumber of defects.

Missions :

• Define Quality criteria: Start by defining the criteria for assessing user story quality. Consider factors like clarity, completeness, correctness, consistency and alignment with projectgoals. These criteria will beused to train our AI model.
• Data collection: Gather a dataset of user stories that have been labeled with quality scores. You can dothis manually by having your team assess the quality of a sample of user stories. The dataset shouldinclude both high-quality and low-quality examples.
• Data preprocessing: Preprocess the text data. This includes tasks like tokenization, removing stop words and converting text to a numerical format that AI models can understand.
• Select an AI model: Choose a machine learning or deep learning model suitable for NLP tasks. Common choices include recurrent neural networks (RNNs), convolutional neural networks(CNNs) or transformer-based models like BERT or GPT.

Date d’expiration: 06 février, 2024