PROFES 2024
Mon 2 - Wed 4 December 2024 Tartu, Estonia

Product development teams often struggle to add value-en- hancing features without increasing maintenance costs at the same time. A data-driven approach, especially through controlled online experiments (A/B tests), is crucial. A/B testing compares a control variant (exist- ing product) with a treatment variant (modified product) in real-world settings, allowing companies to make informed decisions based on user behavior data. This paper explores how AI can streamline the experimen- tation lifecycle by enhancing efficiency and reducing manual workload. Based on a qualitative-empirical study, we identified AI use cases in each step of the lifecycle, which could facilitate the experimentation activities. Focusing on AI’s role in hypothesis formulation, experiment design, and data analysis, the paper advances the understanding of how to automate and optimize experimentation in product development. The presented framework guides practitioners in identifying potential use cases of AI in the product experimentation lifecycle.

Wed 4 Dec

Displayed time zone: Athens change

16:00 - 17:00
PROFES Session 11: AI for SE and Continuous ExperimentationResearch Papers at UT Library - Room 2 (Seminar Room Tõstamaa)
Chair(s): Simone Romano University of Salerno
16:00
18m
Research paper
Insights on the Use of Software Design Principles in Machine Learning Pipelines
Research Papers
Lidia López Universitat Politècnica de Catalunya, Spain, Cristina Gómez Universitat Politècnica de Catalunya, Claudia Ayala Universitat Politècnica de Catalunya, Spain
16:18
18m
Research paper
Use Cases for Artificial Intelligence in the Product Experimentation Lifecycle
Research Papers
Nils Stotz Leuphana University of Lüneburg, Paul Drews Leuphana University of Lüneburg
16:36
24m
Talk
Session 11 Discussion
Research Papers