ADAM: An intelligent virtual assistant for personal financial management

Abstract

Context: According to the Serviço de Proteção ao Crédito (SPC Brasil), approximately 48% of Brazilians do not control their monthly budget. Problem: In addition, according to the Pesquisa de Endividamento e Inadimplência do Consumidor (PEIC), the percentage of families that reported debts reached 76,3% in December 2021. This number indicates the need for solutions to help people manage their personal finances. Solution: This article proposes an intelligent virtual assistant model named ADAM to help manage personal finances. Considering the proliferation and accessibility of instant text messages applications, the model uses chatbots and Natural Language Processing (NLP) to extract information from sentences and texts. NLP allows better user interaction since the users can employ their natural language. IS Theory: The work was carried out following the Design Theory. In addition, the foundations of the Technology Acceptance Model (TAM) were also used mainly to evaluate the artifact. Method: The work adopted Design Science Research as a research method. Problem understanding and the construction of the artifact were carried out with the support of specialists in the finance area. Ten academic community participants and finance specialists evaluated the artifact. Summary of Results: Six users rated the artifact as easy to use and intuitive, facilitating the management of personal finances. Also, suggestions were made to improve the artifact. Contributions and Impact in the IS area: Among the main contributions, the intelligent virtual assistant model stands out, which uses NLP to facilitate interaction with people in managing their finances and bring more autonomy to them.