Product descriptions:
Grupo Casas Bahia is one of Brazil’s largest omnichannel retailers, serving over 100 million customers through more than 1,000 stores and a sweeping national logistics and digital presence. With thousands of customer reviews arriving daily via multiple channels, the company needed a modern, scalable way to analyze feedback and quickly identify opportunities for improvement. By leveraging Databricks Agent Bricks, Grupo Casas Bahia automated the classification of customer comments, delivering actionable insights across departments, boosting analytical efficiency fourteenfold, and ensuring every customer’s voice helps drive the business forward.
Meeting volume and complexity in customer feedback
As Brazil’s retail landscape evolves, customer experience has become a critical differentiator. Grupo Casas Bahia receives a flood of feedback from CES, NPS, Reclame Aqui, App Stores, Procons, and Consumidor.gov, making it challenging to maintain structured analysis and respond to detractor evaluations in a timely manner. “Manual classification was a bottleneck: it took about an hour just to process 100 comments, so we often relied on small samples, leaving gaps in our insights,” explained Brandon Ferreira de Mendonça, Business Analytics Specialist. The process of data extraction and organization was labor-intensive and siloed within a single team, putting business value and agility at risk as the scale of feedback grew.
To unlock the authentic voice of the customer, Casas Bahia needed to automate comment classification, ensuring that every complaint could be categorized by journey, group, and problem — enabling sharper dashboards, faster root-cause analysis, and proactive improvement across digital and physical touchpoints.
Databricks Agent Bricks: Powering a unified, scalable approach
Casas Bahia adopted the Databricks Data Intelligence Platform to consolidate feedback data from multiple internal and external sources in a secure lakehouse, leveraging Unity Catalog for governance and a streamlined data pipeline for ingestion, modeling, and analysis. “The Databricks Platform made it easy to connect with different GenAI models, so we could rapidly test, benchmark, and deploy the best solution for our business,” Brandon shared. The team implemented a custom prompt for Meta’s Llama 3.3 70B model (via Databricks AI Functions), automating end-to-end comment classification and integrating the results into daily dashboards.
The transition from isolated, manual workflows to a collaborative AI-powered process also transformed how data teams operated. Data engineers, scientists, analysts, and user experience specialists now work together seamlessly — enabling flexible experimentation, transparent cost management, and rapid operationalization. “Databricks is always by our side, accelerating our analysis, making governance transparent, and helping us adapt solutions for maximum business impact,” Brandon noted.
Insights, productivity, and customer-centric strategy
The automated comment classification model now processes 33,500 reviews per month — up from just 1,500 with manual sampling. With 90% accuracy in identifying the customer journey tied to each detractor comment, the team quickly pinpoints where experiences break down, tracking 91 distinct problem types across all primary channels. “By automating comment analysis with Agent Bricks, we’ve achieved a 14X productivity gain and an efficiency improvement of more than 9 hours for every 1,000 comments,” said Brandon. “The model saves more than 4,000 person-hours per year — equivalent to nearly R$480,000 — while giving every department a 360-degree, real-time view of customer pain points. This capability helps our business act faster, plan budgets more accurately, and improve satisfaction.”
Corporate dashboards created from this unified data support faster and smarter prioritization of UX, product, and operational initiatives. Stakeholders — from product managers to squad leads and executives — now rely on these insights for strategic planning and to measurably improve satisfaction and engagement. The result is a tighter focus on key improvements, reduced analysis time, and more substantial alignment around what matters most: the customer.
“This transformation not only boosts efficiency but also makes the day-to-day of business teams easier — giving them quicker, more reliable insights that help drive better decisions and customer outcomes,” said João Cuoco, Data Coordinator, Grupo Casas Bahia.
Looking ahead, Casas Bahia plans to expand its analytics model to new feedback channels, refine classifications as language evolves, and monitor how automated insights directly influence NPS and customer experience metrics. “AI has fundamentally changed how we listen and act,” Brandon said. “With Agent Bricks, we can respond to detractor feedback in near real time — accelerating resolutions, improving satisfaction, and building stronger loyalty across every stage of the customer journey.”