Natura &Co deployed a generative AI application in August 2025 that automates gross margin analysis across its Latin American operations, replacing a manual reporting process that had slowed financial decision-making for years. The Brazilian cosmetics group-owner of Natura and Avon-built the tool on SAP Business AI Platform in a six-month co-innovation sprint with SAP and digital transformation partner Numen, and it is now live in the company's Ecuador operations.
The application connects directly to data in SAP S/4HANA and delivers automated insights with narrative recommendations to finance teams in real time. Users can explore revenue, cost, and margin drivers interactively, identifying which elements are protecting or eroding margin performance across markets and product lines. Finance professionals retain full control over interpretation and decisions-the AI generates the insights, but humans decide what to do with them.
"We overcame delays and raised the standard of insights by integrating margin analysis from SAP S/4HANA with a large language model connected via the SAP AI Core layer," said Rogério Dias Garcia, tech manager for ERP Latam at Natura &Co. "This architecture allowed us to provide, in an agile, secure, and completely anonymous manner, a stratified and precise view of gross margin offenders and protectors-discriminating exactly which revenue or cost elements were driving market performance."
Why manual margin analysis broke down
For a company managing a portfolio as diverse as Natura &Co's-spanning multiple brands, product categories, and country markets-gross margin analysis had become a bottleneck. Finance teams spent the bulk of their time pulling data from disparate systems and building offline reports. By the time the analysis reached decision-makers, the numbers were already stale. The delay meant margin erosion often went undetected until quarterly closes made the damage visible.
The new application shifts that dynamic. Instead of requesting reports, finance professionals query the system directly and receive structured answers that break down which cost elements or revenue streams are moving margin in either direction. The underlying architecture pulls from SAP S/4HANA, SAP AI Core, SAP Fiori, and SAP Business Technology Platform, creating a secure, integrated pipeline from transactional data to generative AI output. For finance teams exploring how AI reshapes their function, this kind of embedded intelligence represents a practical step beyond dashboard-based AI for Finance tools.
How the collaboration came together
Numen created the initial prototype during SAP's global Hack2Build on business AI in early 2024. Rather than leaving it as a proof of concept, Natura &Co, Numen, and SAP committed to a production-ready build aligned with concrete business priorities. The six-month sprint focused on scalability from the start, avoiding the trap of building a standalone demo that would never survive contact with real financial data.
"SAP enabled the transformation by providing the technological foundation and expert support," said Carlos Aravechia, head of Data Design & Intelligence at Numen. The project has reinforced a broader conviction inside Natura &Co: that generative AI, embedded directly in ERP workflows, can reposition finance from a transactional function to a strategic business partner. For managers looking to build these capabilities, an AI Learning Path for Finance Managers can provide the foundational knowledge to evaluate similar implementations.
What comes next
Natura &Co is already planning to integrate Joule Agents to automate the extraction of standard analytical content and deepen AI-driven optimization of financial processes. Dias Garcia confirmed the company is "ready to move forward-deepening these insights and integrating the capability of Joule Agents to maximize the extraction of standard content and further optimize our business decisions."
The project also serves as a replicable model for other SAP customers. The formula is straightforward: pick a high-value, well-defined business process, embed AI directly into existing workflows, and build in human oversight from the start. The combination of structured ERP data with the contextual reasoning of large language models creates a foundation for decision intelligence that goes beyond what traditional business intelligence tools can offer.
Why this matters for finance professionals
The Natura &Co case demonstrates that AI in finance is not about replacing analysts-it is about removing the manual data work that keeps them from doing actual analysis. When a system can surface margin offenders and protectors in real time, finance teams shift from reporting what happened last quarter to influencing what happens next. For finance professionals evaluating AI adoption, the takeaway is practical: start with a single, painful reporting process, insist on embedding the AI where the data already lives, and keep humans in the decision loop. The technology is ready. The harder question is whether your reporting workflows are.
Your membership also unlocks:
