PROTEONEXT is a demonstrator platform showing how to combine federated AMR data, generative peptide AI, active learning, and confidential computing on Azure without moving sensitive rows.
PROTEONEXT is an alliance between artificial intelligence and medicine against superbugs. It is a Spanish project seeking CDTI funding to address a global crisis: highly resistant bacteria using artificial intelligence.
Generative AI creates antimicrobial peptides: new weapons against pathogens. Think of a tireless locksmith trying millions of virtual keys until finding the ones that fit the lock of a bacterium that normal antibiotics can no longer open.
With federated learning on Microsoft Azure, AI learns without extracting private hospital data. Like a student entering the library, learning the lesson, and leaving: knowledge travels, but sensitive books never leave the building.
Candidate peptides are synthesized and tested in Petri dishes to see whether they attack bacteria without being toxic. Those results feed back into the AI to improve the next round.
This is not about making a magic pill in two days. The goal is to reduce the massive trial-and-error burden in the lab, moving from trillions of possibilities to the best candidates ready for preclinical testing.
The web app turns the PROTEONEXT memo into a navigable experience for customers, scientific partners, and internal teams: problem, data, federation, AI, simulated lab, and Microsoft governance.