Publication / 2024

Can Large-Language Models Replace Humans in Agile Effort Estimation? Lessons from a Controlled Experiment

A controlled experiment on generative AI for agile effort estimation, examining estimation accuracy and support for planning.

Development teams upgraded an existing information system using either AI-supported estimation or conventional methods. The study reports limited estimation accuracy alongside potential support for task breakdown and iteration planning. Its findings do not support replacing human judgement with the evaluated tool.

AI4SEAgile
Metadata sourceCrossref ↗

Bibliographic details checked against publisher-deposited records; the summary is based on the published abstract.

Related research areas

Related events

· Workshop · OTS 2026

USERSTORY.md

Vasilka Saklamaeva and Lidija Vincekovič’s OTS 2026 workshop on clear requirements, user stories and acceptance criteria.

Past event