With over 18,000 member pharmacies and 65 million customers in its loyalty program, Febrafar manages a massive volume of retail data and marketing activities. One of its most pressing operational challenges was the manual validation of thousands of store photos for trade marketing campaigns. Each image had to be reviewed by the sales team to confirm product placement and compliance, creating delays that could stretch from days to over a week.

"By June 2025, we had received almost 42,000 photos, and around 8,300 were rejected," says Rodrigo Leonavas, head of engineering. "Those rejections were happening manually, which meant a lot of time lost and slower execution for our partners." Scaling the program from one industry partner to the 70+ Febrafar envisioned would've required significant headcount increases, something the organization wanted to avoid.

At the same time, Febrafar was operating across multiple cloud environments. While some products had already been moved to Google Cloud, others remained with a previous provider. Over time, the differences became clear. "Much more than looking at inefficiency from the previous provider, what made us choose to leave a multi-cloud structure was the efficiency and service we found in Google Cloud," says Ney Henrique, director of innovation and technology.

The selection process came down to technology, integration, and trust. Febrafar valued BigQuery's ability to handle large-scale analytics, the availability of AI tools like Gemini and Cloud Vision, and the ease of integrating these services with Cloud SQL and other core systems. Close support from Google's team and partner IPNET further tipped the decision. "Google has everything we need in one environment," says Leonavas. "For me, it was an easy choice, and the partnership has been genuine from the start."

Febrafar's trade marketing program connects industry partners with 18,000 member pharmacies, ensuring products are displayed according to planograms. Until recently, verifying these displays was manual: store owners took photos, sent them through Febrafar's app, and waited for the sales team to review each image. Feedback could take anywhere from three days to more than a week.

The new AI-powered workflow transforms that. When a store manager takes a photo, the app checks for a validation keyword, handwritten or printed, placed next to the display. Using Cloud Vision API, the system confirms the keyword. "Cloud Vision had the best result, with 97% accuracy, even on handwritten keywords," says Leonavas. If the keyword is missing, the app instantly rejects the image, prompting resubmission on the spot.

If detected, the image moves to the next stage: Gemini analyzes the products, checking required items are present, positioned correctly, and in the right quantities. This step replaces hours of manual review. "Even finding parts of the keyword is enough for our algorithm to proceed, which ensures speed without sacrificing accuracy," Leonavas explains.

Now, if a pharmacy in a small town submits an image with a product misplaced, they receive instant feedback instead of waiting days. They can adjust the display, retake the photo, and get approval within minutes. The faster loop improves compliance and builds trust with industry partners, who gain assurance that negotiated displays are executed correctly.

The results are significant. Within weeks, the system automatically blocked 20% of images that would have been rejected manually, saving thousands of minutes of staff time. At scale, with 5,000 photos per month, this efficiency could reclaim more than a month of work per employee each year. Just as important, Febrafar can now expand from one partner to 70+ without adding headcount for validation.

This technology also opens the door to entirely new opportunities. "Before, the industry didn't invest in small and medium-sized pharmacies because they couldn't verify execution," says Henrique. "Now, with AI validation, we give them the confidence to invest in stores they never considered before."

The implementation draws on multiple Google Cloud products beyond Gemini and Cloud Vision, including Cloud Run to host back-end services, Cloud SQL for transactional data, and BigQuery for analytics. Partner IPNET supported the migration and provided proof-of-concept scenarios to validate the approach, ensuring the new system could integrate seamlessly into Febrafar's broader digital ecosystem.

By automating a once-manual process, Febrafar has created a scalable, cost-efficient foundation for trade marketing that benefits both its member pharmacies and the industry partners that support them.

For Febrafar, the launch of its AI-powered validation process is just the beginning. The federation sees this as a foundation for a broader digital transformation, one that will scale automation, open new revenue channels for member pharmacies, and position the organization at the forefront of Brazil's retail pharmacy market.

"Our goal is to increase automation and even approve compliant submissions instantly," says Leonavas. "We're already exploring video-based verification combined with store geolocation, so the process can be even faster and more secure."

Future initiatives include expanding AI adoption across departments, introducing conversational interfaces for store managers, and creating data-driven tools that recommend product mixes based on sales history. "We're shifting our mindset," says Henrique. "Instead of building a product and then adding AI, we now design from the start with AI in mind."