Designing multi-model conversational AI financial systems: understanding sensitive values of women entrepreneurs in Brazil

TL;DR

This paper explores how multi-model conversational AI can be designed to support women entrepreneurs in Brazil by understanding their specific challenges and values, leading to tailored recommendations for microbusiness assessment.

Contribution

It provides insights into the value tensions faced by women entrepreneurs and offers seven design recommendations for AI systems in low-income community contexts.

Findings

Identified value tensions influencing AI design for women entrepreneurs.

Developed seven specific recommendations for AI system design.

Enhanced understanding of microbusiness evaluation needs in low-income settings.

Abstract

Small business owners (SBOs), specially women, face several challenges in everyday life, especially when asking for microcredit loans from financial institutions. Usual difficulties include low credit scores, unbaked situations, outstanding debts, informal employment situations, inability to showcase their payable capacity, and lack of financial guarantor. Moreover, SBOs often need help applying for microcredit loans due to the lack of information on how to proceed. The task of asking for a loan is a complex practice, and asymmetric power relationships might emerge, but that benefits micro-entrepreneurs only sometimes. In this paper, we interviewed 20 women entrepreneurs living in a low-income community in Brazil. We wanted to unveil value tensions derived from this practice that might influence the design of AI technologies for the public. In doing so, we used a conversational system as a probe to understand the opportunities for empowering their practices with the support of AI multimedia conversational systems. We derived seven recommendations for designing AI systems for evaluating micro-business health in low-income communities.