ChatGPT-4o in requirements elicitation: an undergraduate thesis runs the case study the community needed
Geraldo Victor Alves Barbosa and João Victor de Oliveira Matos defended at FCTE/UnB an undergraduate thesis investigating the use of ChatGPT-4o in the requirements elicitation, discovery, and specification process.
Brasília, February 19, 2025
Eliciting requirements is a conversation. And as the conversation gains the possibility of including a language model, a question sets in: what changes in the process? The undergraduate thesis defended in February 2025 by Geraldo Victor Alves Barbosa and João Victor de Oliveira Matos takes the question seriously and runs a concrete case study, under the advising of Prof. George Marsicano Corrêa .
The work investigates the use of ChatGPT-4o in three stages of requirements engineering: elicitation, discovery, and specification. Instead of only producing examples or anecdotal essays, the duo adopts the rigor of a case study, with defined scope and evaluation criteria. What surfaces is useful: where the AI adds speed at no quality cost, where it produces the appearance of progress, and where it must be systematically restrained so as not to distort the real needs of the stakeholder.
Operating at the crossing of requirements engineering and AI , the thesis contributes to a debate the community urgently needs: how to integrate LLMs into consolidated technical processes without losing judgment or accountability over the final decision.
The full text — written in Portuguese — is available at the UnB Undergraduate Theses Digital Library.
About CEDIS: The Center for Studies, Development, and Innovation in Software (CEDIS), linked to the University of Brasília, researches and develops innovative software solutions.