BioTech
Trifon Tsekov
CEO @ 3-Fi Medical
Talk TopicFrom Data Silos to Clinical Evidence: Federated Learning for Medical Software
About the speaker
Trifon Tsekov is CEO of 3-Fi Medical, a former Director of Hardware R&D and radar systems architect with 10+ years in embedded systems, real-time signal processing, and rugged defence hardware. An AI Cluster Bulgaria member, he applies high-reliability engineering to regulated medical AI.
About the talk
Talk DescriptionMedical AI teams need evidence that their software is safe, performs as intended, and creates clinical value - but relevant patient data is often fragmented across institutions. This session introduces federated learning as a practical way to work across approved data environments without centralizing sensitive patient data, connecting the concept to SaMD evidence generation under the EU MDR using plain-language examples. It is intended for founders, researchers, hospitals, and innovation teams working to turn strong research into scalable, trustworthy health products.
Key TakeawaysA plain-language view of why clinical evidence matters for medical software beyond model accuracy, and the practical idea behind federated learning - bringing computation to governed data environments instead of moving sensitive data into one central repository. The session shows how multi-site collaboration can support validation, monitoring, and evidence-generation workflows for SaMD products, and highlights what startups, hospitals, and researchers need to align early: intended use, governance, validation criteria, and regulatory documentation.