Back to projects
Projects

Case 1: Bacen + Reclame Aqui Analysis

Tools
Python · LLMs · dbt
Sector
Financial services

Most financial institutions read complaints as operational noise. I read them as market signal. The Central Bank publishes quarterly a complaint ranking per institution, with standardized classification. Reclame Aqui continuously publishes complaints in free text, written by the customer themselves, identified by institution.

Cross-referencing the two with semantic classification via LLMs, it's possible to map where customer experience drifts from regulatory standards, and where the next competitive opening lives. Not hypothesis: it's what customers are literally saying, at scale.

[Analysis in progress. Each institution below will have a dedicated page with specific reading of what appears in the data.]

Analyses by institution