Click here to read this story in Portuguese

For cancer patients, time is the most precious resource. The difference between early detection and a delayed diagnosis often dictates the course of treatment and the likelihood of survival. Late-stage cancers demand more intensive and expensive interventions, placing immense strain on both patients and healthcare systems. Consider Brazil, a nation of immense scale and genetic diversity, where access to preventive screenings is critically low. With 80% of Brazilian women of screening age skipping mammograms, the sheer volume of undiagnosed cases presents an urgent, systemic problem for public health.

This stark reality inspired Vinicius Ribeiro, Daniella Castro, and Marco Kohara to launch Huna in 2022. The trio of founders had their own deeply personal motivations for building the startup. Daniella saw untapped potential in routine blood test data after a friend’s cancer diagnosis; Vinicius was frustrated by research that never reached patients, and Marco, who lost his mother to cancer, believed AI could bring screening to those most often left behind.

The use of a blood test in cancer screening

Huna’s core idea is both simple and powerful—turn a standard complete blood count (CBC), one of the most common and accessible blood tests, into a scalable cancer screening tool. By training AI models on data from over one million women—linking CBC results to confirm cancer outcomes—Huna uncovered subtle patterns in blood that point to disease, long before symptoms appear. The company partnered with leading healthcare institutions like Hospital de Amor and Grupo Fleury to validate its approach using real-world data.

Empowering Huna's solution with AI

The Huna team leveraged tailored support from Google for Startups AI Academy and Accelerator programs to cut AI model training times by 50%, enabling faster iteration and responses to new clinical needs. Google Cloud credits allowed the team to test their solutions at scale without incurring high infrastructure costs. Experts also helped Huna establish robust data systems and model operations—critical for building dependable AI in healthcare. Leveraging Gemini, Google’s large AI model accessed through Vertex AI on Cloud, Huna was able to automate the complex task of interpreting a range of reports, multiplying the impact of their lean team and enabling them to develop multiple tools in parallel. Open-source frameworks provided by Google further sped up prototyping, helping the company move from concept to deployment quickly and effectively.

"Previously, we faced the challenge of manually labeling exams and creating complex rules for each cancer type. Now, with Gemini, we have achieved equal or even superior accuracy, much faster. This has significantly scaled our ability to simultaneously develop products for multiple types of cancer," states Daniella Castro.

Scientific research as a central point for growth

The impact of this work is already being felt. Huna’s technology has enabled earlier identification of up to two times more breast cancer cases in a group of nearly 400,000 patients. Their growing database includes more than five million unique patient records, covering nearly two decades of clinical history—giving their models a uniquely powerful and representative foundation. A new pilot with the public health system from Brazil’s Northeast region is bringing Huna’s technology to 100,000 patients, demonstrating the company’s potential to drive population-scale change. Huna’s AI identified a high-risk group with the expected likelihood of developing breast cancer, allowing providers to act earlier and with greater precision.

Scientific rigor has been central to Huna’s growth strategy. In May 2024, Nature published a study co-authored by Huna’s founders, which became one of the top 100 most-read articles of the year. Peer-reviewed research like this not only validates Huna’s technology but also builds lasting credibility with healthcare professionals, researchers, and policy leaders. As co-founder Vinicius Ribeiro puts it, “While others focus on high-cost, limited solutions, we’re building tools that labs and hospitals around the world can use right now—using tests they already have.”

Huna’s mission doesn’t stop at breast cancer. The company has already expanded its work to include cervical and colorectal cancer, publishing new studies and presenting results at major medical conferences. Their long-term vision is to support the full cancer care journey—from early detection to personalized treatment—with AI that helps prioritize biopsies, predict recurrence, and optimize therapies.

The Huna team is not only advancing new technology—it is shaping the future of equitable, scalable healthcare. By transforming routine blood tests into early warning tools, Huna is helping to shape a world where life-saving cancer detection is smarter, faster, and available to everyone.