RTU MIREA has launched a new master's degree course titled Cloud Technologies, developed in partnership with the medical company SberHealth. The curriculum focuses on designing secure cloud platforms for medical AI assistants, emphasizing architectural integrity and legal compliance over mere algorithmic capabilities.
Students learn to build systems where AI agents assist therapists by summarizing hospitalizations and analyzing clinical data. The architecture requires deterministic code for fact-checking, mandatory source citation, and human oversight for all final medical decisions. Practical training utilizes synthetic data to ensure patient privacy.
The program covers advanced technical stacks, including model context protocols, orchestration scenarios, and security policies as code. Students are trained to implement safeguards against malicious prompts and memory distortion, while also managing the operational costs and observability of AI agents.
The course concludes with the defense of a functional cloud agent platform prototype. Students must demonstrate system resilience under load and justify their architectural choices, focusing on creating tools that remain under human control within the healthcare sector.



